Social Consequences of the East Asian Financial Crisis€¦ · Social Consequences of the East Asian Financial Crisis. 3. East Asia’s economic and social structures are under strain,

Responding to the Global Financial Crisis

Social Consequencesof the East AsianFinancial CrisisTamar Manuelyan Atinc and Michael Walton

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Pub

lic D

iscl

osur

e A

utho

rized

Administrator

31041

Contents

Context 1

Progress and Vulnerabilities 4Impressive measures of social development 4Pre-crisis challenges and emerging vulnerabilities 7

The Social Impact of the Crisis 14Effects on households 14Future impacts of the crisis on poverty 19

What can be Done 23How economic policies affect the poor 23Food, employment, and income security 27Maintaining economic and social services for the poor 32Institutions, corruption and the social fabric 36The central role of public information 41

Conclusion 43

Bibliography 46

This report was prepared by Tamar Manuelyan Atinç and Michael Walton withinputs from Stephen Commins and Caroline Robb. The authors are grateful toVajeera Dorabawila for assistance, Shaohua Chen and Martin Ravallion for povertyprojections, and Benu Bidani, Kene Ezemenari, Francisco Ferreira, ScottGuggenheim, Emmanuel Jimenez, Katherine Marshall, Andrew Mason, RobertHolzman, Haneen Sayed, K. Subbarao, Gurushri Swamy, Jack van Holst Pellekaan,and participants in seminars in Jakarta and Manila for contributions and comments.The paper is a slightly extended version of Chapter 5 of East Asia: the Road toRecovery, World Bank, 1998; it was originally prepared for the Asia DevelopmentForum in March 1998 in Manila.

Social Consequences of the East Asian Financial Crisis1

Context

This paper is a slightly extended version of the essay that forms chapter 5 in theWorld Bank’s report, East Asia: The Road to Recovery.1 It is being producedas a separate, free standing paper, in order to provide wider access to the WorldBank’s views on the linkages between the financial crisis and the social crisis,in East Asia and elsewhere. It is in particular intended to serve as one of severalways in which the World Bank is attempting to expand its engagement withvarious institutions and partners, including civil society organizations,concerned with the human impact of the crisis.

Amidst the turmoil in global markets, the human toll of financial crises shouldremain at the core of policy decisions and responses for countries in economiccrisis. At the time of writing, financial upheavals are spreading with alarmingspeed, with much of East Asia and Russia already engulfed in financial crisis,Latin American under threat, and at least a risk of global recession.2 The humancosts of such a scenario could be greater than the setbacks of the decade thatfollowed the eruption of the debt crisis in Mexico in 1982.

The response to the social fallout from the 1980s debt crisis can best becharacterized as too little, too late. The World Bank was one of many culpableparties to this weak response3. The introduction of measures to mitigate adversesocial effects into adjustment programs only became standard years after thecrisis started, and then often in the form of add-ons to structural adjustmentprograms. By contrast, in the crisis that started in East Asia in 1997, socialissues have been placed at the forefront of public and international concern.However, it remains an extraordinary challenge to frame responses that willmoderate the social cost of macroeconomic crisis. The effects of economy-widedecline are so pervasive that public action to shore up services and offsetincome declines for the poor and those newly vulnerable due to economicupheavals is often at best partially effective. Moreover two central areas ofimpact are still only weakly understood: the distributional consequences ofchoices over macroeconomic and structural policies; and the inter-relationshipsbetween economic shocks and the social fabric, including patterns of socialinclusion and exclusion, of the performance of local institutions and of informalcommunity-level coping mechanisms. Yet it is vital that the economic andinstitutional reforms being designed in the crisis countries serve to mend andrepair the social fabric in those countries and support equitable growth whenthe recovery occurs. Separating economic policy making and socialdevelopment is a recipe for failure.

The present crisis started in the region that was the paragon of povertyreduction over the past few decades. East Asian societies have experienced apace of poverty decline, in terms of income aspects of living, but also health andeducation, that is almost certainly unprecedented in human history, and farquicker than the pace of progress on poverty experienced by the now-industrialized countries in their periods of rapid take-off. They have beenpraised by many both within East Asia and outside (from Oxfam to the World

The World Bank Group2

Bank) as models of human development for others to emulate, because of theinclusive, equitable pattern of development and the priority to provisioning ofsome basic social services. However, while the gains were genuine, it is notacceptable to return to a "status quo ante" in the crisis countries. Even beforethe crisis, there were concerns over gaps and weaknesses in social systems,including concerns over rising inequality, inadequate mechanisms for socialprotection and weak and corrupt public service institutions, often linked toauthoritarian social structures. These have been sharply exacerbated by thecrisis. Recognizing the gaps that were pre-existent to the crisis provides thebasis for developing longer term priorities for establishing new foundations forsocial development and new institutions for social protection against futuredownturns.

The scope and scale of the tasks in the crisis countries is more than can bemanaged by the governments and people in each country, requiring theinvolvement of external partners seeking to work towards common goals ofrenewed growth and poverty reduction. The linkage between economic andinstitutional reforms on the one hand, and reducing poverty on the other,requires deeper and more effective engagement between the World Bank andvarious partners, including the diverse organizations of civil society. This paperis designed to present some of the World Bank's views on aspects of the socialcrisis. It is not intended to be the final word but rather an invitation to furtherdialogue and consultation at the national and regional level. The World Bank isseeking to work with governments, research organizations and civil society toassess the emerging social impacts, to raise questions about policies andproposed projects, and to work with governments, civil society and otherexternal agents to assess the shifting pattern of vulnerability and theconsequences of both the economy-wide shocks and responses to inform andshape future public action. A number of initiatives have been established thatare outlined at the end of this paper. The core of the paper that follows exploresboth the longer-term issues and the crisis for the main affected countries.

East Asia in Crisis: What does this mean for the poor?“It was the rich who benefited from the boom…but we, the poor, pay theprice of the crisis. Even our limited access to schools and health is nowbeginning to disappear. We fear for our children’s future” said KhunBunjan, a community leader from the slums of Khon Kaen, North EastThailand and her husband, Khun Wichai. Khun Wichai recently lost his jobin the local factory and his wife is selling less at the local market. As aresult they took both their son and daughter out of school and put them towork. “What is the justice in having to send our children to the garbage siteevery day to support the family?” questions Khun Bunjan. But Khun Wichaithinks he is lucky. His neighbors are sending their children to beg and somegirls became prostitutes. Among the older male youths, drug dealing hasbecome an increasingly attractive source of income. Increased competitionfor survival, frustration and psychological stress are all leading toheightened household and community tension. This tension has led toincreased domestic violence and with fewer jobs, neighbors who oncecooperated are now competing. Stealing, crime and violence, are on therise. People are feeling unsafe and insecure. “This breakdown of ourcommunity's networks will affect stability,” added Khun Bunjan.


East Asia’s economic and social structures are under strain, and decades ofunparalleled social progress are under threat. In country after country, growthis declining from an average per capita increase of over 5 percent per annum tonegative levels for the next year or so with sharp contractions expected inIndonesia, Thailand and Korea. The poor are being severely hurt during thecrisis as demand for their labor falls, prices for essential commodities rise,social services are cut, and crop failures occur in countries experiencingdrought. The combined macroeconomic and agricultural shock undercuts fragilecoping mechanisms, especially in Indonesia, and could be life-threatening ifconsumption levels drop sharply. Widespread economic hardships are tearing atthe fabric of society: there are food riots and ethnic tensions in Indonesia,farmers are protesting in Thailand, and workers are expressing their discontentin Korea. These signs of social stress are not only politically worrisome but theincreasing tensions at the household, and community level are equallydamaging, especially for women and children. Children are being pulled out ofschool and put to work; food is being rationed within the household, and womenand girls are frequently the first to sacrifice their portions; and violence, streetchildren, and prostitution are all on the increase. This is the human crisis.

A speedy return to macroeconomic stability and growth through distributionallyfavorable adjustment policies is the only way to begin to put a floor under thefalling incomes of the poor. In the meantime, public action can curb the welfarelosses of the poor in the short-term, and help protect their human resourceinvestments. It is essential to ensure that food markets work, to augment thepurchasing power of vulnerable households, to cushion the impact of priceincreases, and to preserve the poor’s access to health and education.Strengthening public and private institutions responsible for service delivery isalso crucial in both the short and longer-term.

East Asian countries encountered important social challenges even before thecrisis. In most of East Asia, social policies developed against the backdrop ofpolitical stability, full employment, high household savings, and relativelystrong community ties giving governments little reason to plan for downsiderisks. But even during times of economic growth, three issues were alreadychallenging social policies in East Asia: persistent pockets of poverty and risinginequality, outmoded labor market policies and industrial relations, and risingneeds for formal mechanisms to support household security. Growth maskedthose problems but when the crisis stripped this mask away, the region’spersistent social vulnerabilities were sharply revealed.

Progress and Vulnerabilities


ast Asian countries have achieved spectacular welfare gains in the last twodecades.4 Consistently high growth rates have been translated intoquantifiable welfare improvements primarily because growth has largely

been inclusive. Public provisioning of social services has been widespread andthe productivity of the poor and their employment opportunities have increasedenormously. The number of poor has fallen and the severity of poverty hasdeclined. Life expectancy at birth, infant mortality, and literacy have allimproved. These achievements are even more impressive when compared withsocial developments in other regions or developed countries during their decadesof industrialization.

Impressive measures of social developmentBetween 1975 and 1995 poverty in East Asia dropped by two-thirds accordingto the region’s head-count index with respect to the constant US$1-a-daypoverty line (in 1985 purchasing power parity terms), and the pace of povertyreduction was faster than in any other developing region. In 1975, six out of tenEast Asians lived in absolute poverty according to this standard; by 1995, theratio had dropped to two out of ten (see table 1). This means that the number ofpoor people in the region was more than halved, from 720 million to 345million. Further, the rate of decline accelerated after 1985. The number ofpeople in poverty fell by 27 percent in 1975–85; in 1985-95 the decline was 34percent.

Changes in poverty levels and rates varied across the region. In 1975, 92percent of the region’s poor lived in China and Indonesia, primarily becausethey were the two most populous countries. Since then, however, both countrieshave seen substantial declines in poverty: 82 percent in Indonesia and 63percent in China. In absolute terms, the number of poor was more than halvedin China, and fell by almost three-fourths in Indonesia. As a result, by 1995 thetwo countries’ share of the region’s poor had dropped to 84 percent. Andalthough Indonesia’s record is remarkable—the head-count declined from 64percent in 1975 to 11 percent in 1995 (and further by 1997see below) -Thailand had the largest proportional reduction between 1975 and 1995, from 8percent to less than 1 percent.

E


Table 1: Poverty in East Asia, Summary Statistics: 1975–95Economy Number of people in poverty

(million)Head-count Index

(percent)Poverty Gap

(percent)1975 1985 1995 1975 1985 1995 1975 1985 1995

East Asia 716.8 524.2 345.7 57.7 37.3 21.2 n.a. 10.9 6.4

East Asia (exc. China) 147.9 125.9 76.4 51.4 35.6 18.2 n.a. 11.1 4.6

Malaysia 2.1 1.7 0.9 17.4 10.8 4.3 5.4 2.5 <1.0

Thailand 3.4 5.1 <0.5 8.1 10.0 <1.0 1.2 1.5 <1.0

Indonesia 87.2 52.8 21.9 64.3 32.2 11.4 23.7 8.5 1.7

China 568.9b 398.3 269.3 59.5b 37.9 22.2 n.a. 10.9 7.0

Philippines 15.4 17.7 17.6 35.7 32.4 25.5 10.6 9.2 6.5

Papua New Guinea n.a. 0.5 1.0c n.a. 15.7 21.7c n.a. 3.7 5.6c

Lao PDRd n.a. 2.2 2.0 n.a. 61.1 41.4 n.a. 18.0 9.5

Vietnam n.a. 44.3e 31.3 n.a. 74.0e 42.2 n.a. 28.0e 11.9

Mongolia n.a. 1.6 1.9 n.a. 85.0 81.4 n.a. 42.5 38.6

n.a. not available.Notes: All numbers in this table (except for Lao People’s Democratic Republic) are based on the international poverty line of US$1per person per day at 1985 prices.a.: includes only those countries presented in the table.b.: Data relates to 1978 and applies to rural China only (World Bank 1996d).c.: Data relates to 1996.d.: Available data on PPP exchange rates and various price deflators for Lao People’s Democratic Republic are not very reliable andlead to anomalous results. The poverty numbers for Lao People’s Democratic Republic in this table are based on a national povertyline which is based on the level of food consumption that yields an energy level of 2,100 calories per person per day and a non-foodcomponent equivalent to the value of non-food spending by households which are just capable of meeting their food requirements. TheUS$1-a-day poverty line is based on characteristic poverty lines in low income countries that have comparable basis in food and non-food consumption needs; the poverty numbers for Lao People’s Democratic Republic are, therefore, not strictly comparable to thosefor other countries.e.: The figures refer to 1984. “Vietnam Household Welfare in Vietnam’s Transition” in Macroeconomic Reform and PovertyReduction, edited by D. Dollar, J. Litvack, and P. Glewwe. World Bank Regional and Sectoral Study, 1998.

Source: Everyone’s Miracle?, World Bank 1997.

Between 1973 and 1990, the region saw substantial gains in life expectancy anddeclines in infant mortality (see table 2). Similarly, access to educationexpanded—China and Indonesia reportedly joined Korea and the Philippines inachieving universal primary net enrollments. In all four countries, and inMalaysia, secondary net enrollment also expanded beyond 50 percent ofchildren in the eligible age group.


Table 2: Social Indicators in East Asia, 1985-95Country Life expectancy at

birthInfant mortality rate

(per 1,000 livebirths)a

Primary netenrollment(percent)

Secondary netenrollment(percent)

1970 1995 1970 1995 1970 1995 1970 1995East Asia 59.4 68.8 76 34 na Na na naTaiwan, Chinaa 69.0 74.8 69 6 na > 99 75.0 87.4Korea 60.6 72.0 46 10 > 99 > 99 45.4 93.4Malaysia 61.6 71.8 45 12 84.1 88.7 25.5 55.9Thailand 58.4 69.0 73 35 78.6 88.2 18.2 34.9Indonesia 47.9 63.7 118 51 75.6 > 99 13.0 55.0China 61.7 69.4 69 34 75.9 > 99 34.7 50.7Philippines 57.2 66.5 71 39 > 99 > 99 40.4 75.5Papua NewGuinea

46.7 58.5 112 68 30.8 70.0 3.7 13.3

Lao PDR 40.4 52.8 146 104 na 60.0b na 15.0b, c

Vietnam 49.36 67.5 111 42 na 91.0d na 45.0c, d

Mongolia 52.70 66.4 102 55 na na na Na

Notes: na. Not available A: Source: Republic of China (ROC), various years. b: 1993, Source: World Bank, 1995a. c: lower secondary. d: Source: World Bank, 1996e.Source: Ahuja and Filmer (1996), for net enrollment rates; World Bank data for life expectancy and infant mortality.

Five factors contributed to the region’s social progress and inclusivedevelopment.5 Some could be jeopardized by the crisis.

• Smallholder-based rural development. In most East Asian countries,small-scale family farming has dominated agricultural production, andgovernment policies have contributed to an equitable development path bysupporting productivity growth on family farms through infrastructure(notably in irrigation and roads), appropriate pricing and other marketpolicies, plus new technologies.

• Rapid growth in demand for non-agricultural labor. In all of the region’smajor economies, except the Philippines, the past few decades havewitnessed a massive labor shift—out of agriculture into more productivework such as rural non-farm activities, urban industry, and services. Thisshift was driven by many factors: agricultural incomes rose, spurringdemand for rural non-farm employment, internal economic integrationincreased, overall capital deepened, and industrial employment grewrapidly. In Malaysia the share of wage workers in industry and servicesrose from 30 percent in 1960 to more than 60 percent in the 1990s.6 Sincethe early 1980s, rural, and then urban industry and service employment hasexpanded enormously in China. Such changes have been accompanied bylarge increases in real wages—and by large increases in the incomes of self-employed workers.


• Widespread public provision of basic education and health services.Rapid growth in the public provision of schooling was a major element ofthe region’s human resource strategy. Primary schooling expanded rapidly,followed by growth in secondary, and then tertiary education. These effortscontributed to the early achievement of almost universal primary education,complemented by a significant expansion of basic health services, includingkey preventive services such as immunizations and basic curative care.

• ”Flexible” labor markets and low labor market dualism. East Asian labormarkets are fairly flexible, with fewer institutional or policy-drivenrigidities than European or Latin American markets—minimum wagepolicies are limited; wage-setting practices are flexible, and wages andproductivity growth are closely linked. As a result, fewer sharp contrastsexisted between formal, privileged workers and rural, informal workers.

• Upgrading workforce skills and investing in education ahead of demand.In Korea, education expansion effectively anticipated the changing demandsof modern industries and services. Young people with a primary schooleducation were the core of the workforce for the early phases of labor-intensive industrialization. As productivity and wages rose—driven by highlevels of capital investment and technological advance—demands shiftedfirst to secondary and increasingly to tertiary graduates. Meanwhile,schooling expanded rapidly enough that skills supplies actually surpasseddemands. Between the early 1970s and the late 1980s, this prompted adecline in wage inequality among workers with college, secondary, andprimary schooling.7 While young school graduates tended to enter andremain in industry, older workers were moving out of agriculture directlyinto services. Rural productivity rose rapidly and the unskilled labor markettightened, prompting a convergence of rural and urban incomes.8

But not everything was fine, even in these domains. We turn next to the majorpre-crisis issues. For example, Thailand’s sharp rise in inequality appears to bepartly related to slow growth in secondary education; China’s unusually smalltertiary education sector could cause future problems. In Indonesia and thePhilippines there are concerns about education quality. Some East Asiancountries have done very well in terms of health, but by some measuresunusually high levels of child mortality exist in Indonesia, Korea and thePhilippines.9 In some cases, labor market policies have been overly restrictive,especially in Korea and China, where insiders are highly protected and stateintervention in employment is pervasive. Measures to tackle insecurity aremixed—Korea is following Japan toward a European style, pay-as-you-go,welfare state; China is providing cradle-to-grave protection for state sectoremployees and their families; Malaysia and Singapore are running large-scale,state-managed provident funds.

Pre-crisis challenges and emerging vulnerabilitiesEven before the onset of the crisis, three major vulnerabilities already existed,but they were masked by widespread growth: protracted poverty and risinginequality; concerns about labor rights; and rising demands for formalmechanisms to offset household insecurity. The crisis has aggravated conditions


underlying each of these issues. These each have to be addressed in order tobuild the foundations for a caring and competitive society. The crisis has nowmade facing these issues imperative for all the countries in crisis.

Persistent poverty and vulnerabilityDespite tremendous gains, poverty remains a fact of life for many in East Asia.Poverty is still high in Indochina and Mongolia, reflecting slower growth andrecent systemic transition. But even in high growth countries, vulnerabilitiesremain considerable given the large numbers of households just above thepoverty line: in Indonesia, a 25 percent increase in the poverty line results inmore than doubling the headcount index, from 11 to 25 percent in 1996.Moreover, absolute poverty persists in certain areas or among certain groups.The poor tend to live in rural areas, have less education and live in householdsheaded by farmers. In addition, some ethnic minorities are disproportionatelypoor and girls appear to get shortchanged in household resource allocation—particularly in poor households.10

Reductions were not uniform across these economies, and poverty remainsacute in some regions. For example, in 1990 poverty incidence in Indonesiaranged from 1.3 percent in Jakarta, to 46 percent in West Nusa Tenggara. Inthe inland province of Guizhou in China, the incidence of poverty in 1992 was20 times that in the booming coastal province of Guangdong. In Thailand, thenortheast has the highest incidence of poverty and the highest concentration ofpoor people. Even in Vietnam and Lao PDR, where poverty is much morewidespread, regional variation is substantial. In Vietnam the incidence ofpoverty ranges from 34 percent in the southeast to 77 percent in the north-central region.

Are some regions poor because they contain a high concentration of householdswith characteristics that are strong correlates of poverty? Or are there purelygeographic effects? Jalan and Ravallion (1997) suggest that in Chinageographic effects remain strong even after controlling for householdcharacteristics: households with identical characteristics experience different oreven diverging consumption growth depending on location.11 This findingsuggests that policies to augment geographic and community capital areessential to help alleviate poverty.

Rising inequalityEast Asia has been credited with achieving “growth with equity” but the factsare more varied. On average, East Asia’s income distribution has remainedlargely unchanged in the last 15 years, either in absolute terms or relative toother regions. An aggregate index of inequality finds that East Asia is moreegalitarian than Latin America or Sub-Saharan Africa but less egalitarian thanformer socialist countries in Eastern Europe, high-income countries, and SouthAsia (see table 3).


Table 3: An International Comparison of Inequality(averaged Gini coefficients)

Region 1980s 1990s

Eastern Europe 25.0 28.9

High-income countries 33.2 33.8

South Asia 35.0 31.9

East Asia and the Pacific 38.7 38.1

Middle East and North Africa 40.5 38.0

Sub-Saharan Africa 43.7 47.0

Latin America and the Caribbean 49.8 49.3

Note: The total sample includes 108 economies. Although Gini coefficients come from household surveys that satisfycomprehensiveness criteria in terms of both geographical coverage and income sources, they nevertheless includeunadjusted data from both expenditure and income distributions. The proportion of income Gini coefficients vary acrossregions, hampering comparability. Regional averages are unweighted, and changes across the two decades may be dueto changes in the composition of the sample. The numbers merely suggest broad orders of magnitude.

Source: Deininger and Squire (1996)

But there was significant variation among countries (see table 4).12 Inequalityhas clearly risen in China, Hong Kong (China), and Thailand. It also appears tohave inched up in the Philippines between 1985 and 1994, and recent data for1997 suggest a sharper increase.13 Only Malaysia shows a slight decline ininequality, although this reflects earlier gains which were partially reversed by asignificant rise in inequality in the 1990s. With the exception of China, all fourcountries now have inequality rates well above the regional average.

Research on China and Thailand suggests two explanations for risinginequality. First, the returns to higher levels of education have increased, whichis driving a wedge between highly skilled workers and those with primary orlower-secondary education. Second, spatial disparity in economic prosperity isgrowing because activity is concentrated in certain areas. The current crisismay increase inequality in access to education and also affect regionsdifferently, further aggravating skills- and geography-based differentials.

Table 4: Inequality in East Asia


Gini coefficient(percentage points)

Economy Period Measure variable First year Last year

Hong Kong 1971-91 I/H 40.9 45.0

Singapore 1973-89 I/H 41.0 39.0

Taiwan, China 1985-95 I/P 29.0 31.7

Korea, Rep. Of 1970-88 I/H 33.3 33.6

Malaysia 1973-95 I/P 50.1 48.5

Thailanda 1975-92 E/P 36.4 46.2

Indonesia 1970-95 E/P 34.9 34.2

Chinab 1985-95 I/P 29.9 38.8

Philippines 1985-94 E/P 41.0 42.9

Papua New Guinea 1996 E/P 50.9

Lao PDR 1993 E/P 30.4

Vietnam 1993 E/P 35.4

Mongolia 1995 E/P 33.2

Note: I/P is per capita income, E/P is per capita expenditure, and I/H is income per household. The numbers in this table maybe marginally different than those reported in other World Bank reports based on unit record data. For the sake ofconsistency across countries, we only report Ginis based on grouped data, except for Korea, Singapore, and Hong Kong,(China) which are from Deininger and Squire 1996. a. Thailand is the only country for which we can present Ginis based on both expenditure and income distribution. Theper capita income-based Gini (I/P) was 42.6 percent in 1975 and 54.6 percent in 1992. b. Because of China’s size as well as with valuing home production of grain for own consumption, controlling for spatialprice variations, and valuing in-kind transfers, the uncertainty associated with Chinese Ginis may be even greater thanthat for other economies (see World Bank 1997b for a detailed discussion).

Source: Deininger and Squire (1996), and World Bank staff calculations.

High inequality is bad for society for three reasons: it undermines povertyalleviation, it may impede growth, and it can contribute to social tension. For agiven growth rate, an increase in inequality tends to lead to slower povertyreduction.14 Greater initial inequality can reduce economic growth because ofimperfect credit markets or through political economy channels. Social tensioncan result when the benefits of growth accrue unequally to easily identifiablegroups—for example, certain regions, ethnic groups, or men and boys—even ifthese are not major factors in overall inequality. Japan, Korea, and China arenotably homogenous societies, but Indonesia and Malaysia have major dividesacross ethnic lines that have spilled over into significant conflicts in the past—into the blood bath of 1965 in Indonesia and periodic violence against Chinese-Indonesians since, notably in 1998. Yet, both achieved sufficient socialunderstanding and stability to foster high levels of capital investment, especiallyby ethnic Chinese groups. In Malaysia’s case there was a highly managedprocess of affirmative action for the ethnic Malay population, or Bumiputra(see box 1). Managing inequality is a challenge both for rich and poorcountries, because it is not easy to balance incentives for superior individual


performance with acceptable levels of inequality which foster poverty reductionand growth.

Box 1: Malaysia’s NEP and social inequality

Ethnic rioting came to a head in Kuala Lumpur in May 1969 after a Chinese-dominated opposition won many parliamentary seats from the Malay-dominatedruling coalition. Malays, most of whom were impoverished, reacted violently infear of losing their political influence. To confront and combat the causes forthis uprising, the government launched the New Economic Policy (NEP), anaffirmative action plan designed to pull poor Malays back into the mainstreamof the country’s economic system.

The NEP introduced a series of government regulations, quotas, scholarships,and other privileges designed to help Malays. The results have been impressive:Malay’s share of national wealth jumped to 20.6 percent in 1995 from 2.3percent in 1970. Much of the NEP’s success is attributed to education as thenumber of Malay doctors, lawyers, and engineers increased dramatically.Malaysia’s affirmative action policies have not been without controversy andsome have observed that the policies helped foster favoritism and inefficiency.Even so, the overall approach clearly succeeded in supporting high levels ofinvestment, low levels of internal conflict, and rapid advancement of allMalaysians, especially the Bumiputra, for more than two and a half decades.

Source: “Far Eastern Economic Review”, 5/28/98; “Lessons From Malaysia,”by Murray Heibert.

Labor relationscountries in the region sought to maintain relatively unfettered labor marketsduring the early stages of economic development at the expense of grantingworkers’ rights to bargain collectively. However, the apparent flexibility was inpart a product of high growth and the large supplies of labor in agriculture (orabroad for some countries). There was more underlying segmentation than inthe common characterization, and also emerging problems of wagesoutstripping productivity growth in countries with tight labor markets (notablyKorea). In some cases highly regulated labor markets developed in the modernsector.15 Until the late 1980s, labor conditions in the five East Asianeconomies—Korea, Malaysia, the Philippines, Taiwan (China), and Thailand—were determined unilaterally by employers with or without the assistance ofgovernment (Galenson 1992). However, in the late 1980s changes began toaffect unions. There were shifts towards more democratic governments,pressures for modern industrial relations brought about by tighter labormarkets, and more sophisticated production processes.

Failure to modernize worker–management relations in countries withsophisticated economic and political structures can become costly, as Korea’s


experience since the late 1980s has shown. If there is no industrial relationssystem to allow workers to air grievances and resolve disputes, strikes and otherforms of job actions, sometimes violent, can become common. One of thechallenges facing maturing economies in East Asia is that of managingindustrial relations to protect workers’ legitimate rights, but avoid grantingentitlements that result in resource inefficiencies.

Household risksMost East Asian households have few formal mechanisms to protect them fromrisks associated with job losses, disabilities, and aging. Instead, most relyprimarily on personal savings and informal family and community links. A fewcountries in the region have set up formal schemes to address householdinsecurity, but these cover small portions of the population (state workers inChina, large enterprises in Korea, participants in the state provident funds inMalaysia and Singapore)the only cases where coverage is substantial. Thecurrent financial crisis makes glaringly evident the absence of formalprovisioning for household security for many households.

In the face of the crisis, the demands for safety nets are urgent but they will befar more dramatic 25 years from now because East Asian societies areundergoing a rapid demographic shift. East Asian populations are aging,moving into cities, and increasingly working in the formal sector (see figure 1).In France it took 140 years for the proportion of the population over 60 todouble—from 9 to 18 percent. By contrast, in Korea, people over 60 willdouble their share in the population in only 30 years—between 1990 and2020—and in China the portion of people over 60 will rise from 9 percent to 16percent of the population during the same period.16 These changes reflect aswifter demographic transition, and in China, they demonstrate the combinedeffects of early gains in rural health status and an activist population policy. Allthree trends will strain informal family-based mechanisms of householdprotection, and will increase demands for formal, government-mandatedschemes. ��


Figure 1: Urbanization, Aging, and Labor

Force Adjustments in East Asia East Asia’s population is urbanizing ...

0.0

10.0

20.0

30.0

40.0

50.0

60.0

1990 2020

3HUFHQWRIWKH3RSXODWLRQ/LYLQJLQ

8UEDQ

$UHDV

with China

w/o China

.. greying rapidly ..

0 2 4 6 8 1 0 1 2 1 4 1 6

C hina

In d on es ia

K orea

M a lays ia

P erce nt of the popu la tio n o ve r 6 0 y ea rs o f age

199 0

202 0

.. and increasingly working in the formalsector as incomes rise.

0

20

40

60

80

100

0 5000 10000 15000 20000 25000

PPP GNP Per Capita

3HUFHQWRISRSXODWLRQLQ

LQGXVWU\

DQGVHUYLFHV

ZDJHHP

SOR\P

HQW

China

Hong Kong

Indonesia

Philippines

Thailand

Malaysia

Korea


The Social Impact of the Crisis

he economic contraction is affecting the lives of millions, and aggravatingsocial vulnerabilities. It is likely to have many dimensions—fallingincomes, rising absolute poverty and malnutrition, declining public

services, threats to educational and health status, increased pressure on women,and increased crime and violence. In Indonesia there is also a radical breakdownin social order as an increasingly fragile social equilibrium was brought underintolerable stress by the collapse in economic confidence and fall in incomes.

The effects of the crisis are acute in Indonesia, and severe in Thailand, Korea,and Malaysia. The Philippines has been less affected, but also shows signs ofworsening social conditions. (After declining steadily for five years, inSeptember 1997 there was a rise in self-reported poverty).17 Trade, capitalflow, and migration linkages among countries are hastening the transmission ofeconomic and social effects across the region. While China remains largelyinsulated, falling regional demand and slowing intra-regional foreign investmentare aggravating domestic difficulties. Countries in Indochina are experiencinggrowth downturns and financial difficulties as the impact of the regional crisisunfolds with a depth and intensity far exceeding previous expectations. ThePacific Islands have also been hit; by far the worst affected is the SolomonIslands where GDP is expected to shrink by 10-12 percent, driven by acollapsing log export market in which export prices have halved.

Effect on householdsThe economic crisis is hitting households hard in four ways: falling labordemand, sharp price shifts, a public spending squeeze, and erosion of the socialfabric. In addition some countries have been simultaneously hit by drought.

Falling labor demand. Economic decline, the corporate crisis, and a creditsqueeze are causing lay-offs, real wage declines, weak demand for new labormarket entrants, and falling margins in the informal sector. Whether the impactin a given country is primarily through higher unemployment or lower wagesdepends on societal and economic structures. In Thailand, unemployment hasincreased by 50 percent since the start of the crisis to 1.5 million (February1998) and is expected to exceed 6 percent by year-end. In Korea,unemployment reached 7.0 percent in June 1998 and may affect as many as 2million people during the year, up from 0.5 million last year (see figure 2). Inthe Philippines, 1 million additional people joined the ranks of the joblessbetween April 1997 and April 1998, raising the unemployment rate to 13.3percent. In Indonesia, where some 4.5 million people (4.9 percent) were alreadyunemployed in 1996, official estimates suggest an additional 10 million maylose their jobs by early 1999, although it is likely that many of them will moveinto low-paying urban and rural informal sector work rather than into openunemployment.

T


Figure 2

Unemployment is on the rise Joblessness in the Philippines ishigh and edging upward..

.. while in Korea it is lower butsharply on the rise.

0

2

4

6

8

10

12

14-DQ��

$SU��

-XO��

2FW��

-DQ��

$SU��

-XO��

2FW��

-DQ��

percent

0

1

2

3

4

5

6

7

-DQ��

$SU��

-XO��

2FW��

-DQ��

$SU��

percent

Source: Korean and Philippine Authorities

This is not essentially an urban shock, despite the high profile of urbanunemployment figures. Rural areas will be affected by labor movements,production linkages, and intra-household relationships because of the highlyintegrated nature of the economies and the declining demand in urban areas.Increased under-employment and falling wages may be more widespread andvalid indicators of a decline in well-being than open unemployment. In Korea,open unemployment has increased sharply, but family and agricultural workersand labor force drop-outs are also on the rise suggesting substantial increases inunderemployment. Data from Western Java (Indonesia) point to a decline inrural real wages by 10 percent between August and December 1997. Availableinformation suggests that workers are returning to their villages from Jakartaand Bangkok. There is an international dimension as well: migrant workers inMalaysia, notably from Indonesia, are also losing their jobs or suffering realwage declines. The Philippines is also vulnerable owing to large numbers ofoverseas workers. Finally, there is evidence that women are disproportionatelytargeted for lay-offs as societies regress to traditional perceptions of men as the“primary bread-winners” (see box 2) .


Box 2: In Bad Times, Women Suffer the Most

One key area that is essential for policy making and about which too little isknown is the gender aspects of the crisis. There are various anecdotal reportson the burdens that have fallen on women in different countries, but evidenceis difficult to find. Given existing research on gender dynamics, it is likelythat a combination of social factors, cultural views and changing economicopportunities mean that the financial upheaval has had different impacts onwomen than on men. As in most of the world, there are long-term biasesagainst women in East Asia: they are generally less powerful than men athome, at work and in politics. Nevertheless, women shared in the massiveabsolute economic and social gains during the decades of growth, and evenenjoyed some reduction in relative disparities, for example in the labormarket.18 But women and girls may be disproportionately hurt by thefinancial crisis. Women lose their jobs first, and families pull their daughtersout of school before sons interference. The particularly hard pressed may selldaughters to brothels. Even before the crisis, girls in Indonesia were six timesmore likely than boys to drop out of school before the fourth grade. Once girlsare removed from school, they rarely go back.

When income shortfalls require reductions in food intake, women and girlssometimes face disproportionate cuts. Social organizations also point to a risein domestic violence and prostitution. Though gender inequality is not a newproblem in the region, the crisis situation has exacerbated the difficulties facedby poor women and girls in East Asia.

To better understand the gender dynamics in the different countries, andbetween different sectors within these countries, action oriented researchneeds to be undertaken on issues such as trends in formal employment,changes in educational opportunities, and access to health services. The socialcost is often carried by women through reduced food intake at home, or worsethrough domestic violence that reflects the growing erosion of social capital.Research is especially important at a time of shifting economic relations, aswomen may find themselves facing unemployment or multiple employment.Anecdotal stories have also presented dramatic images of children who havedropped out of school and who have out of necessity taken to the streets.While programs have been instituted, as outlined in this paper, to keepchildren in school, the aggregate and disaggregated impacts are unknown.Wildly disparate numbers are cited in terms of the percentage or number ofchildren who have fallen back in to poverty. In order to understand what ishappening in the lives of children, new research programs are needed that willprovide a clearer and more up to date information. One example is that theparticular situation of adolescent girls becomes more uncertain whenemployment opportunities or further educational programs are cut off, leadingto pressure for early marriage, exploitation or harsh working situations. Yet,little is known beyond a few stories about children in garbage dumps and girlsforced into prostitution----stories that present stark images of the social costs,but which cannot adequately provide for longer term programs or new policyemphases.


Sharp price shifts. Prices have risen because of exchange rate devaluation,associated increases in the prices of traded goods, and public sector price hikes.In Indonesia, where the exchange rate has depreciated by 80 percent since July1997, the prices of antibiotics doubled between October 1997 and March 1998and the consumer price index (CPI) for food increased by more than 50 percentbetween June 1997 and March 1998 compared to a 38 percent rise for thegeneral CPI. Prices of drugs have risen sharply throughout the region—dramatically in Indonesia, where there are already reports of householdspostponing vaccinations or use of other drugs; HIV/AIDS patients in the regionmay be particularly vulnerable. The longer-term effect of these changes isdifficult to predict because in most countries the exchange rate appears to haveovershot an equilibrium position, and it is hard to know how much will beinflated away or reversed by increased demand for local assets once confidencerecovers. But there will clearly be short-run price effects. These will squeezeprofits in much of the informal sector, that is generally already facing weakdemand in the domestic market.

Public spending squeeze. Public spending is being constrained by revenue falls,the effects of exchange rate changes on interest bills and the need to finance theincreasing liabilities of the government in financial sector and corporaterestructuring. Budgets were cut initially in all of the affected countries as partof the macroeconomic adjustment process. While fiscal targets have since beenadjusted in some countries, public services may still suffer cutbacks creatingboth short-and-long-run impacts on households. Of particular concern arepotentially irreversible effects on investments in human resources. Constraintsmay also mean funding is cut in areas that have a serious impact on relativelypoor households (such as smallholder treecrop investment).

Erosion of the social fabric. Economic stresses are leading to social andpolitical problems. Rapid development has brought rising incomes to most EastAsian households in the past decades. It has also led to rapid social change,urbanization, migration, and expansion of schooling. A sudden stop to the rapidgrowth can be expected to disturb the social equilibrium. Social unrest inIndonesia vividly showed how fragile this equilibrium can be, and how quicklysocially repressive societies can be thrown into upheaval once civil societybegins to question its leaders’ abilities to deliver continued improvements ineconomic well-being—the central source of political legitimacy. At a lessdramatic but nevertheless significant scale, social stresses are mounting at boththe household and community levels in all crisis countries. Focus groupdiscussions suggest that households with falling incomes are coping byincreasing the workload of mothers and by taking children out of school andputting them to work.19 Economic stresses may also be leading to increaseddomestic and community violence and resorting to illegal activities such asprostitution and drug trade.

Drought. Parts of Indonesia, the Philippines, and Thailand have been hit bydrought due to El Niño. While the effects appear to be less adverse thanoriginally feared, in some areas the combined macroeconomic and agriculturalshock makes coping particularly difficult. Consumption drops can be life-


threatening, especially in Indonesia where distribution channels and marketshave been damaged (in part by ethnically focused violence) and foods stockshave dwindled. In the Philippines, the large increase in unemployment betweenApril 1997 and April 1998 was entirely due to the sharp drop in agriculturalemployment as a result of El Niño, particularly among unpaid family workersand independent farmers. Households that have suffered from repeated droughtsfor the past several years—for example, in Indonesia’s Eastern Islands—areparticularly vulnerable as they are likely to have run down their assets andstocks to smooth consumption. The extent of crop failure varies widely andfood price rises are magnifying these distributional effects. Higher prices helpfarm households which can raise surplus food to sell, but hurt long-term fooddeficit households and farm households temporarily in food deficit because ofthe drought.

These shocks will affect incomes, well-being, access to services, and willinteract with the coping strategies that households adopt to protect theirconsumption levels (see box 3). The effects on the poor will depend on the depthand duration of the macroeconomic recession and on whether distributionworsens or improves during the crisis. Both dimensions are uncertain. Otherepisodes of economic contraction provide limited insights into expected trendsin distribution. In many Latin American countries, distribution worsened duringthe economic difficulties of the 1980s, but in Malaysia’s most recentcontraction in 1984–87, distribution improved enough to prevent a rise inpoverty levels.20


Box 3: People are Struggling to Get By

Entering the informal labor market. Household income falls in all countrieshave already forced many families to send more women, children, and elderlyinto the labor force. In Thailand, NGOs report an increase in child labor, childprostitution, and child beggars. In the slum settlement of Teparak, Khon Kaen,northeast Thailand, women were angry because they had to send their childrento the garbage site every day to support the family. In Indonesia, there arereports of children leaving school to join padat karya programs (labor-intensiveprojects).

Migrating. In Thailand, rising urban unemployment is likely to force industrialworkers back to their villages of origin. A Tambon representative, in the ruralvillage of Sap poo pan, northeast Thailand, estimated that out of the villagepopulation of 260, 40 people had already returned because of the crisis and 70were still working outside the village, mainly in Bangkok. Returning laborerswill increase the competition for jobs, further marginalizing farm laborers whoare less competitive. The impact of migration from neighboring countries is stilldifficult to define. In Cambodia’s northwest province, whole communities reliedon jobs in Thailand. Focus groups reported that within the last three months,most of these workers are returning because they have lost their jobs and arestill in debt.

Cutting down on household expenditure. In slum areas, people reported cuttingdown from three meals per day to two, or even one. In Indonesia, (Maluku andSouth Sulawesi) school principals complained that parents were havingdifficulty paying parent association fees on time, or at all. In Indonesia and thePhilippines (Mindanao), teachers reported that children were eating less beforecoming to school in the mornings and buying less from vendors and that thiswas affecting students’ ability to concentrate.

Turning to illegal activities. In some slum areas in the Philippines andThailand, people report an increase in crime. The slum dwellers in Bangkok toldfocus groups that unemployed youth were already turning to selling drugs tosupport their families. And, in Cambodia, there were reports of increases in thetrafficking of women and children.

Source: Rapid social assessments undertaken by Bank staff in collaborationwith NGOs in Thailand, Philippines, Indonesia and Cambodia, January-April,1998.

Future impacts of the crisis on povertyForecasting poverty is hazardous enough in conditions of stability. So, wesimply illustrate potential outcomes for the poor of alternative assumptions ongrowth and distribution for four affected Southeast Asian economies.21 Results


are shown based on two commonly used poverty lines: US$1 per person per day(in 1985 purchasing power prices), which is close to poverty lines used inpoorer countries (and close to the Indonesia poverty line) and US$2 a person aday, which is closer to national lines used in middle-income countries. InMalaysia and Thailand there are few people living on less than US$1 a day, butmore than 15–20 percent survive on less than US$2 a day. By contrast, inIndonesia and the Philippines half the population still lives on less than US$2 aday.

Poverty is highly sensitive to both growth and changes in inequality; the impactalso depends on the initial pattern of distribution. This is explored for four ofthe crisis countries by examining the consequences of common assumptions foreach of them (that is, this is not a forecast). First, we stimulate the effect onpoverty of a 10 percent fall in aggregate consumption or income between 1997and 2000 using a poverty line of US$1 a day for Indonesia and the Philippines,and US$2 a day for Malaysia and Thailand. With no change in distribution,poverty would double in Indonesia, and increase by 35 to 50 percent for thePhilippines, Thailand, and Malaysia. If, in addition, inequality worsened, therewould be a large additional effect on poverty. A 10 percent rise or fall in theGini coefficient—a significant change by past standards—in the distribution ofincome illustrates the possible orders of magnitude, and would have a majorinfluence on projected poverty outcomes (see figure 3). In Indonesia, forexample, a 10 percent worsening in inequality would cause poverty incidence toalmost triple—from less than 7 percent (estimated for 1997) to almost 20percent in 2000. By contrast, if inequality improved by 10 percent, povertywould remain largely unchanged.

We then illustrate the impact on poverty of the most likely growth scenarios foreach country for the period between 1998 and 2000 (based on consensusforecasts) with no change in inequality. Country-specific annual growth ratesare used for this exercise to calculate poverty incidence in the year 2000 (seefigure 3). For Indonesia, the large expected declines in aggregate growth pushan additional 10 percent of the population into poverty by 2000 even if there isno worsening in inequality. This is equivalent to a rise in the number of peopleliving under US$1 a day from 13 million (estimated for 1997) to 34 million for1999 and 2000compared with 22 million in 1995 and 53 million in 1985(table 1). Thailand is also expected to be hit hard with a 20 percent increase inthe head count index. Malaysia and the Philippines are projected to show mildereffects although simulations point to a deceleration of gains in the Philippines,and a reversal of a long-term trend of poverty alleviation in Malaysia. Anyworsening in the distribution would aggravate poverty outcomes.


Figure 3: Poverty projections (poverty incidence, percent) A 10 percent contraction in aggregate GDP between 1998 and 2000 would add millions to the ranks ofthe poor. Beyond growth, distributional changes also strongly influence poverty outcomes. But evenwith no change in inequality, the most likely growth scenarios (1988-2000) imply a halt or a sharpreversal in a 20- year long trend in poverty alleviation.

Indonesia ($1 a day)

1997

2000 -10% growth

10% increase in inequality

10% decrease in inequality

6.6

19.6

13.2

6.0

No change in inequality

-15.0% growth

No change in inequality16.3

Philippines ($1 a day)

1997

2000-10% growth

21.5

34.1

29.1

22.6

20.3





8.7 % growth

M alaysia ($2 a day)

1997

2000-10% gro w th

19.9

31.4

25.7

21.8





0.3 % grow th

18.3

Thailand ($2 a day)

19 9 7

2 0 00-1 0 % g ro w th

14.7

27.5

20.2

10.5

N o c h an g e in ine q ua lity

N o c h an g e in ine q ua lity17.6-5 .5 % g ro w th

1 0% inc re as e in in eq u ality

1 0% d e cre a se in in e qu a lity

Note: The growth rates used for simulating poverty incidence refer to cumulative rates over the three-year period,1998-2000. Source: Ravallion and Chen 1998.

The scenarios show the powerful influence of overall economy-wideperformance on poverty through its impact on both average incomes andinequality. But what can be said about actual short-run distributionaloutcomes?

It is difficult to draw general lessons from past episodes, since there are nogeneral patterns to distributional changes in periods of contraction. But, it ispossible to explore the relationships between short-run sectoral growthexpectations and the structure of poverty. This has been done for Indonesia byapplying projected sectoral growth rates to the structure of household incomesavailable from the consumption survey.22 The simulations assume majordeclines in construction (-35 percent), commerce (-18 percent) and financialservices (-18 percent), and a drought-induced slowdown in agriculture (1percent), yielding a 12 percent decline in aggregate GDP. They examine only


impact effects, and do not take account of either household responses orsecond-round effects, notably in the labor market.

In Indonesia, a 12 percent decline in aggregate GDP in 1998 is estimated toincrease poverty from 11 percent in 1996 to 14 percent in 1999, affecting bothrural and urban residents. Even if second-round effects are ignored, mostpoverty in Indonesia would remain rural. Yet the immediate effect of the crisisis expected to be more acute in urban areas, resulting in a significant scale ofurban poverty—8 percent in 1999—for the first time in many years (see table6). Some groups would be particularly hard hit: the construction industrycollapse would push poverty incidence among workers’ families from 8 percentto nearly 31 percent, but the actual number of people involved is relativelysmall. Of course, workers and households will respond and move into otherlabor markets. Java in particular has a high degree of inter-connectednessbetween rural and urban areas. If agricultural incomes hold up, as the scenariosnow indicate, there could be a rise in rural inequality: those with land wouldgain in relative terms, as those without land suffer from the general collapse inlabor demand.

Table 6: Impact effects on poverty in Indonesia(percentage of population)

Baseline

January 1996

Forecast 12 percent contraction

March 1999

Rural 15.0 17.6

Urban 5.0 8.3

Total 11.3 14.1

What is the cost of eliminating the crisis-induced increase in poverty throughpublic transfers? We use Indonesia to illustrate magnitudes.23 In Indonesia, theincrease in the poverty gap between 1996 and fiscal year 1988–89 is estimatedto be about 1.2 trillion rupiahs, or about 1 percent of 1988–89 budgetaryrevenues (see table 7). Restoring the poor’s consumption to its 1996 levelswould require a transfer of this amount. There are many ways this could bedone, ranging from a universal (that is, untargeted) cash transfer to varioustargeted schemes, including some commodity subsidies and employmentprograms. An untargeted cash transfer, assuming no administrative expenses,would cost about 8 trillion rupiahs; the increased costs reflect the substantialleakages to the non-poor. But is possible to do better. Analysis done forIndonesia puts the cost of transferring US$1 through public employment (padatkarya) programs at US$3.8-US$5.3; transfers through such programs wouldthus require 4.6–6.5 billion rupiahs or 3.5–5 percent of budgetary revenues.(see table 7) These are small amounts (as they usually are), especially whencompared to the costs of financial sector restructuring, but administrative


difficulties can be daunting and the actual costs of effecting such transferswould be very much greater.

Table 7: Cost of selected transfers to eliminate crisis-induced poverty inIndonesia(in current rupiahs)

Increase in poverty gap between 1996 and

1998/99

1.2 trillion

Cost of closing the gap through

Universal cash transfersa 8 trillion

Padat karya-type programs 4.6-6.5 trillion

Notes: a Assuming no administrative cost


What can be Done?

There is an urgent need to reduce crisis-induced welfare declines. Much can belearned from successes–and failures–in responding to past crises, notably inLatin America (see box 4). Public action can make a difference in six domains:

• The design of economy-wide policies

• The policy and institutional framework for labor markets and incomesecurity

• The level and pattern of economic and social services

• The interrelationship between corruption, institutions and the social fabric

• The central role of public information.

How economic policies affect the poorAll the East Asian crisis countries are experiencing sharp drops in domesticdemand. This is especially true of investments which are expected to fall by 10percentage points of GDP per year, or more. This raises two questions: willfuture poverty reduction be jeopardized by the loss of productive investmentand would compensating expansionary policies help the poor?


Box 4: What can East Asia learn from the experience of Latin America?

Latin America has experienced two major crises in the last 15 years; both began inMexico and both incurred substantial social costs. In the 1980s, Mexico's debt crisisspread throughout Latin America on the back of trade shocks and weak publicfinances. In 1995, Mexico's liquidity crisis, with its roots in excessive private lending,spread only to Argentina.

Latin America's two crises hit households hard. In the 1980s, real wages in Argentinaand Mexico fell by nearly 40 percent, while poverty increased by more than 30percent. In Chile, real wages shrank by about 15 percent and unemployment rose by 9percentage points within a year. In the 1995–1996 crisis, real wages in Mexico fell bymore than 30 percent. In Argentina, unemployment rose by 6 percentage points andremained at around 18 percent of the work force for more than two years. As a result,poverty increased by more than 50 percent. Latin American governments respondedwith a variety of programs, but with the partial exception of Chile, the overallresponse was too little, too late. What are the lessons?

Solving social problems means putting them at the top of the agenda. Too often in theLatin American crises, policy makers' energy was devoted to restoringmacroeconomic stability and implementing structural reform. East Asia has theopportunity to avoid this mistake by putting social issues at the forefront. Drops inincome, employment, and public services have widespread and complex socialconsequences. Therefore, it is important to take action on a wide range of fronts, andmake every effort to anticipate these consequences.

Targeting is crucial. Special funds set up in Latin America, designed to cushion thesocial costs of economic adjustment measures, had mixed success, largely due to poortargeting. Chile’s large public works program was successful, in part because it setwages low enough to ensure that the most needy were the main participants.

Cutting subsidies especially of goods that are essential items to the poor can fuelresistance to needed reforms, unless effectively targeted alternatives are offered. Thiswas the case in the Mexican crisis.

Maintain support for core education and health services, or risk irreversible losses inhuman investment.

Social funds may be useful alternatives to government safety net programs, but onlywhen the Government really lacks the cohesion or resources to make programs work.Poor targeting, lack of genuine community participation, and long set-up timeplagued many of the funds established in Latin America. There is now moreknowledge on how to design such funds well. But if established governmental ornongovernmental institutions work reasonably well, as they did in the past in most ofEast Asia, it is often better to work within that system.

More democratic systems do not hinder reforms. The Latin American debt crisisoccurred as democratic rule was being restored in many countries and there werefears that democracy would either hinder painful but necessary reforms, or that thesocial disruption and violence associated with the crisis would hasten the return ofauthoritarian rule. Both fears proved groundless. Although the process of adjustmentwas slowed in some of the countries by granting an independent voice to unions,legislatures and other actors, the policy measures that were finally undertaken workedprecisely because they had been endorsed by a more open civil society.


Source: International Herald Tribune, 5/29/98, “East Asia Can Learn From LatinAmerica’s Travails”, by Nora Lustig and Michael Walton.

Protect investments that help the poor. An ‘investment pause,’ is a rationalresponse to a shock. In times of significant drops in incomes, it makes sense tosmooth consumption over time—allowing more consumption now, at the cost ofless in the future. This is of particular value to the poor, or those at risk of largedrops in income. Indeed, a pause in investment helps explain why thecontractions in Indonesia and Malaysia in 1984–87 occurred without any rise inconsumption-based poverty.24 However, there is a cost in lost future growth inincomes–high levels of investment were at the center of the growth process thatbrought such extraordinary poverty reduction in the past. While someinvestments, such as property development, were unproductive, on the wholeEast Asian countries put their investments to good use.25

Different investment categories have different effects on the poor. To reducepoverty it is particularly important to invest in the capabilities of poor children(that if foregone may be irreversible), economic infrastructure in rural andperiurban areas, and private investment in labor-demanding activities. The poorhave a strong interest in overall policies that provide the confidence andfinancial sector preconditions for investment, especially in labor-intensivemanufacturing and services. They also benefit from protecting public spendingfor economic and social services that reach the poor, either directly asbeneficiaries, or via effects on labor demand during the construction phase.

More expansionary macroeconomic policy can help the poor, but there aretradeoffs. To help reduce poverty there is a strong case to err on the side ofexpansion. What scale of fiscal expansion is desirable still involves assessingother intertemporal tradeoffs since a more expansionary position could reduceinvestor confidence, or risk higher inflation and lower future growth prospects.Moderating economic contraction is important for the poor, and may bedistributionally favorable.26 How the fiscal expansion is financed is equallyimportant. Deficit financing would be particularly damaging as there isevidence from the Philippines in the early-1980s, and Brazil from 1980 to theearly 1990s, that the poor are the hardest hit by high inflation.27

Interest rate policies will have mixed effects on the poor. High interest rates canmoderate inflation, restore capital inflows and encourage more labor-intensivenew investment, all potentially pro-poor. But in East Asia in 1998, thedominant effects are the short-run consequences of high interest rates on output,and consequently on labor demand, that is central to the welfare of the poor.

Real exchange rate depreciations will help the poor in the long-run but short-runmeasures are important to cushion the effects of price increases. In theorthodox analysis, real exchange rate devaluation helps the poor because itfavors agriculture—the most severe poverty is rural—and (following a periodof adjustment) it encourages more labor-intensive investment choices. Theevidence from the Philippines from the early 1980s, found that on balance realexchange rate depreciation tended to help the poor. However, these effects are


hard to generalize, and price effects are likely to be highly diverse acrossdifferent groups in the population. Some of the poor are likely to benefit: ricefarmers in Thailand, smallholder export crop producers in the Philippines andIndonesia. But many poor households are net consumers of tradable goods. Thisapplies to most in the informal sector. The poorest groups in both Indonesia andthe Philippines are rural wage earners. Where drought has hit, many morefarmers are net-consumers of food than in a normal year. All of these groupsare vulnerable to the price shock: the next section discusses the role of measureson the pricing and labor demand side to offset these effects.

Financial sector bailouts should not spare equity-holders. Financial sectorpolicy matters for the poor; the management of bailouts and effects onmicrocredit are of particular concern. Dealing efficiently with rotten parts of thefinancial system will require substantial infusions of public funds—perhaps onthe order of 30 percent of GDP. While this is manageable within prudent debtlimits, owing to low initial levels of public debt, it diverts vast resources fromalternative uses to finance unsound or even corrupt past lending. The loans aremostly gone—whether to real estate follies or foreign bank accounts—and it isstandard practice to protect small depositors who are mainly middle income;few among the poor have significant levels of deposits. However, protectinglarger depositors to avoid a collapse in confidence is costly. At a minimum,distributional and political economy considerations demand that owners ofequity in the financial sector take the hit before taxpayers. This is also moreefficient since it reduces incentives for owners to go for high-risk lendingstrategies in the future.

East Asian countries have a wide range of financial institutions operating inmost villages. Indonesia’s BRI Kupedes is one of the world’s largest ruralmicro-savings and credit programs. The Philippines has a wide array of small-scale programs, and there are significant, if smaller, sectors in Malaysia andThailand. While many of these cater to better-off rural dwellers, they aresometimes an important part of the village economy. There have been concernsover the consequences of the crisis for the micro-finance movement. A recentsurvey of micro-finance institutions found that deposits continued to rise inmany, possibly because they were sounder institutions and rural savers wereshifting out of smaller rural banks.28 Some institutions enjoyed sustained highrepayment rates, others suffered sharp rises in defaults. Efforts should be madeto avoid adverse effects of selective deposit guarantees on rural financialservices. While short-term measures may be needed to avoid the collapse ofmicro-finance institutions, sustainability should remain an important medium-term objective.

Invest in institutional transparency and accountability and in the skills ofpoor people to increase benefits from globalization and temper inequality inthe longer term. The crisis has brought distributional struggles into the open.Three areas are likely to be important to tackle longer-term distributionalconcerns. First, a major societal concern is unfair or corrupt gains among thewealthy and connected. Fostering sound corporate governance structures, openand accountable public sector institutions, competitive procurement and bidding


practices will all help, but this is a domain where the issues are often deeplylinked to political structures and not amenable to technocractic solutions.Second, there is some evidence that the forces of global integration andtechnological advance target the skilled, leading to widening wage differencesbetween those with and without education. 29 The appropriate response is not towithdraw from international engagement—countries that do this hurt themselvesand their poor in particular—but to support inclusive and high qualityeducational systems. Third, in the past growth largely bypassed populationsliving in poor areas. Policies that augment not only human capital, but alsocommunity capital, in these areas, may be helpful but in some casesoutmigration may be the only effective strategy—as in the series of projects ininner China being supported by the World Bank.30

Food, employment, and income securityThe economic crisis threatens the livelihood of the poor. When growth resumes,income gains and employment opportunities will expand, but this may takeseveral years and there may be other temporary pauses in future growth. In theshort term, the poor must be protected against drastic declines in consumption.Measures should aim at ensuring food security and preserving the purchasingpower of vulnerable households. In the long-term, households will needassistance attaining income security during old age, and health and employmentshocks, and to reform labor market policies.

Ensuring Food SecurityFood prices have increased sharply in most countries and there is evidence offood shortages in parts of Indonesia. Price increases on rice primarily affect thepoor. In Indonesia and the Philippines, rice accounts for nearly 20 percent ofhousehold expenditures for the poorest one-fourth of households. In Indonesia,it jumps to 30 percent of household expenditures for the poorest decile. Arelative increase in rice prices could hurt some of the long-and short-term poorand the non-poor hit by drops in labor demand; but it would protect householdsand villages that remain in food surplus.

Managing rice supplies to smooth price adjustments and avoid major pricehikes may be desirable in the short-term—especially in light of the substantialovershooting of the exchange rate. This could moderate short-term inflationaryand political shocks, and would help smooth consumption of poor, food-deficithouseholds in rural and urban areas. But it is costly. In Indonesia, the cost oftransferring one dollar to the poorest 15 percent of households through the ricesubsidy is estimated at US$8.20—no better than a universal cash transfer.31 Ifsubsidies were limited to low-quality foods, the poor would benefit and ricetransfer costs would be reduced to US$3.60 in Indonesia. Beyond the shortterm, there is a case for exploring additional mechanisms for targeting foodsubsidies—for example, through food stamps. While shifting from generalizedto targeted subsidies is good policy in general, the Latin American experienceshows that it is best not done during times of crisis.

Policies to keep food affordable are only useful if food is readily available.Ensuring that food markets function is fundamental to managing the effects of


the drought. But, where markets are breaking down—as in parts of Indonesia—there is no alternative to direct food distribution. A combination of directdelivery to villages, NGOs, or religious organizations, food-for-work programs,and vulnerable group feeding may be desirable.

Sustaining the purchasing power of vulnerable householdsPoor households are losing purchasing power due to falling labor demand andrising relative prices of some commodities. The case for rice subsidies wasdiscussed above. Kerosene is the most relevant subsidized fuel for the poor andnear poor in most of the region. Evidence from 1990 data for Indonesiaindicates that kerosene subsidies are badly targeted to the poor, with anincidence roughly in line with overall income distribution. Some phasing ofprice increases may be justified on political grounds, but they should notcompete with programs targeting the poor.

There are two possible approaches to addressing employment and income lossesdue to falling labor demand: stimulating demand for labor through labor-intensive public works programs, or providing income support for theunemployed. The question is how to target scarce resources and reduce leakagesdue to poor administration, graft, and access by non-targeted groups.32 Acombination of targeting techniques will need to be used, including geographic,household-based, and self-targeting.

Promoting public works. If designed properly, employment programs can helpsustain the purchasing power of households in the short-term, and address long-term issues of declines in seasonal labor demand, as seen in parts of Indonesia.Indonesia is already expanding padat karya schemes and new programs arebeing launched in Thailand; there is also a long tradition of public employmentin the Philippines.

Effective targeting depends on the wage. To ensure that the neediest choose toparticipate and to generate information on where demand for low-wage work isgreatest, local market or lower wages should be used.33 In Indonesia, forexample, ongoing public works programs use a standard Rp7500 wage (in early1998) that is often above local market wages, especially in depressed labormarkets. Under these conditions, there is a greater risk that poor households willbe rationed out of the programs. Programs with a high share of wage costs inthe total will also be more effective at reaching the poor. Finally, it is not clearwhether women have full access to manual work in East Asian public worksschemes. This is in contrast to many South Asian schemes, which often havehigh female participation. These concerns emphasize the need to build in anoverall monitoring and evaluation system and to conduct an ongoing assessmentof short-term transfer and targeting objectives.

Public employment programs have dual objectives—to transfer income andimprove infrastructure. Chile in the early 1980s, and Korea in the late 1980s,are examples of public works programs that were allowed to fade away asdemand fell.34 Indonesia also has a history of using food and cash for work


during periods of adverse shocks, then phasing them out. Several programdesign issues are relevant:

• Balance urban and rural programs. Both programs make sense but anurban bias misses the majority of the poor and may also increase commutermigration, or decrease return migration to rural areas, slowing efficientresponses to the shock.

• Balance cash transfer and productivity. In the short term, cash transfersachieve the primary objective—to increase security. But over time,programs may become more focused on infrastructure production ormaintenance.

• Balance between male and female beneficiaries. Programs should be opento participation by women but special efforts may be needed to publicizepublic works programs, review contracting procedures, and monitornumber of jobs given to them. In addition, certain design features—forexample, piece rate payments instead of daily rates so women can work ingroups and organize childcare—can encourage greater participation bywomen.

• Workfare for the non-poor does not alleviate poverty. Workfare for laid-off factory workers or young work force entrants with secondary or highereducation who may choose not to take traditional manual work has littlerationale on poverty grounds.

Some individuals and households, especially the disabled and householdswithout adult labor, will fall through the safety net of public employmentprograms. In the short term, existing community channels, including religiousorganizations or NGOs, are probably the best source of support for thesegroups.

Unemployment benefits. With the exception of Korea, countries in the region donot have unemployment insurance schemes which can help sustain individualstemporarily out of work. But this is not the best time to introduce insurance-type unemployment benefits schemes in the crisis-affected countries: benefitscannot be awarded immediately as insurance schemes generally requireminimum periods of prior contribution and a waiting period before benefits canbe paid; the introduction of payroll taxes and/or individual contributions wouldincrease the cost of labor (or the reservation wage) and hence depress labordemand (supply); new entrants to the labor market who are unemployed wouldnot qualify; and the scheme would largely benefit formal sector employees.Such schemes can be introduced upon a return to more stable economicconditions, but even then attention should be paid to design features that reducelabor market distortions (see box 5).


Box 5: Unemployment Insurance Lessons of International Experience:Minimizing Labor Market Distortions

Unemployment insurance schemes are not effective at addressing structuralunemployment issues or large labor demand shocks. They are more suited toconditions of frictional or cyclical unemployment which afford possibilities forpooling risks across individuals or time. Even then, it is important to minimizelabor market distortions by:

Observing strict eligibility criteria—prior contribution record, involuntarydismissal, proven willingness to work, acceptable travel time and regionalmobility, etc.

Creating a governance structure to prevent politically-driven generosity inbenefit levels and raiding of the reserve fund during good times

Limiting benefit duration (say, a maximum of 6 months) and referral to socialassistance schemes thereafter

Risk-rating unemployment insurance contributions by economic sector,enterprises, or both, thereby allowing contribution rates to vary as a function ofpast unemployment rates.

In the short-term, a system of unemployment assistance benefits may be moreappropriate. It would aim at alleviating poverty among the unemployed byproviding, for a limited time, limited income transfers in the form of flatbenefits at near the poverty line. This scheme would have a better distributionalimpact than an insurance scheme and would also improve efficiency, as the lowbenefit level would motivate an early return to the labor market and have littleimpact on the reservation wage. Initially, this benefit could be provided to theunemployed without any means-testing. Later this feature could be adjusted orsupplemented by means-tested family, health, and educational allowances.Unemployment assistance benefits can begin disbursing as soon as they areintroduced and would be less demanding to administer than an insurancescheme. Nonetheless, targeting is a concern, especially if informal sectorworkers are to be included, and decentralized mechanisms for administration,including the use of NGOs, will need to be explored.

Training programs also provide some income support during the period ofprofessional reorientation and skill acquisition, even though this is not central totheir objectives. However, this is a costly means of transferring incomes withunclear distributional effects.

Reforming labor market policiesThe question of international competitiveness may seem moot after the largeexchange rate depreciations. But these exchange rate movements have beendriven by capital flows and are likely to reverse once confidence is restored. The


more fundamental issue is productivity growth which is linked to the workforceskill level. The education system must support increasing access to secondaryeducation, but labor market policies are also important:

• Encourage private job placement services to supplement inadequatecapacity in the public sector. There has been rapid expansion of publicemployment offices in Korea, from 53 in 1997 to 113 in 1998 with plans toreach 162 in 1999. This risks undermining the role of private employmentservices. In 1997 private employment services placed 1.8 million workerscompared to 0.3 million by public employment offices. Most workersassisted by private employment offices (83 percent) are engaged in basicand service occupations that usually attract less privileged workers.

• Exercise caution in launching or expanding active labor marketprograms. International experience shows that these programs have littleeffect on employment probabilities or on future wage profiles. Someprograms may be effective for particular categories of workers. Programsshould undergo rigorous monitoring and evaluation and respond to feedbackon effectiveness.

• Change the government’s role in vocational education and training. Theprivate sector can also provide both pre- and in-service training. Increasedcompetition among training providers should increase quality, and thegovernment can ensure regulatory oversight and increased choice forconsumers—through vouchers, for example.

One of the most controversial issues in the labor market is the role of unions.This is because most East Asian countries have restrictive policies toward them,and there is considerable government involvement in industrial–labor relations.Korea was particularly repressive in the pre-democracy period, and reaped abitter harvest in antagonistic labor relations. Its experience carries lessons forother societies.

A key challenge for most East Asian societies is designing a regulatory andinstitutional framework that provides strong rights for freedom of associationand avoids granting entitlements that can cause resource inefficiencies orcrippling inflexibilities. This framework is likely to mean a more neutral role forgovernments—essentially setting the rules of the game. It should also encouragethe primary locus of union activity to be in the enterprise in which collectiveaction is most likely to help increase productivity and less likely to commandmonopoly wage increases.35 There is a role for union confederations whereinformation or dialogue is useful at a sectoral or national level. Developingeffective mechanisms for industrial relations is complex, and workers andemployers must collaborate, often with state support, especially to design asound and neutral regulatory framework.

Managing household insecurity in the longer-termAs the financial crisis demonstrates, temporary downturns in economic activitycan lower living standards for most households. Unemployment, disability, andold age contribute to poverty in both industrial and developing countries, andworsen poverty among people who are already poor.36 In most societies, coping


with insecurity involves some combination of private savings, informal support,and employer obligations. Governments step in when these prove insufficient.Households may find it difficult to borrow to cover temporary drops in income;community support is ineffective when there is an economy-wide shock; andprivate markets for unemployment, disability insurance, and old-age pensionsare often limited or absent.

East Asian countries have relied on informal income security systems for mosthouseholds, especially the poorest. Private transfers help reduce inequality byproviding old-age support, and alleviating the effects of disability, illness, andunemployment. Increasing urbanization and the growing importance of formalemployment have eroded the informal support mechanisms. In addition, there isrising demand for social insurance to deal with household insecurity. A rapidlyaging population further strains family-based support. These trends, plus risingincomes, suggest a much greater role for formal insurance in the future.

What kind of model should East Asian society follow to balance caring andcompetitiveness? The welfare state, created in Europe, is under attack becausegenerous benefits and high taxes are associated with lack of competitiveness,slow employment growth, weak incentives for work, and high unemployment. InEuropean and Central Asian transition economies, universal cradle-to-gravesecurity is waning and in China, the system for state workers is in need ofurgent reform. In many South Asian and Latin American developing countries,the model of employment protection for formal workers is also under attackbecause job security regulations appear to protect insiders at the expense ofcomprehensive unemployment insurance, which encourages the growth ofinformal employment.

Countries in the region have an opportunity to develop schemes that avoid thelabor market rigidities, inequities, and fiscal problems associated with somemodels of social security, and they can learn from the abundant array ofreforms being undertaken worldwide. The central principle is to minimizeadverse labor market and fiscal effects by linking contributions to benefits.Most old-age pensions, certain health risks, and short-term unemployment canbe covered under publicly-mandated insurance and savings schemes thatembody this design feature, but matching contributions directly to benefits isnot always feasible, or desirable. For example, most societies choose to providefor the poorest people far in excess of their potential to contribute—so theprotection schemes incorporate an element of income redistribution. Socializedpublic action could be considered for core pensions for the long-term poor, tocover catastrophic health risks, and to provide social assistance for thedestitute.

Maintaining economic and social services for the poorDuring the crisis, the poor stand to suffer the most—especially from irreversiblelosses in potential education and health that will impede their participation infuture recovery. Efforts to maintain purchasing power will help, but additionalmeasures are needed to focus on keeping schools and health care affordable forpoor households and quality of services intact.


EducationStudies find that public spending on primary schools benefits the poor. Beyondthe primary level, incidence depends on the coverage of the education system. Inlow-and middle-income countries, spending on junior secondary schools is oftendistributed roughly in line with income. Spending on senior secondary andtertiary education tends to be unequally distributed, and can be even moreunequally distributed than income, as is the case in Indonesia.37 However,marginal changes in spending for junior secondary schools—and in richersocieties, senior secondary schools—are likely to benefit the poor. Moreover,analysis has found that in the late 1980s, cuts in education spending wereassociated with significant drops in secondary enrollment However, primaryenrollment remained virtually universal. Since the current economic crisis looksmuch worse than the late-1980s slowdown, there is a strong chance thatchildren in the region will be pulled out of school because of the actual andopportunity costs of schooling. Focus group results from Indonesia and thePhilippines show that this is already happening. Also, there is a risk that forsome children—especially of poorer households—this shift will be permanent.This suggests the need to:

• Preserve real spending on primary schools, and seek to maintain non-salaryspending. In past episodes of adjustment in other countries non-salaryspending has been the most vulnerable to cuts during a fiscal squeeze, withpotentially high costs in quality.

• Increase targeted subsidies to encourage students to stay in secondaryschool, linked as closely as possible to income level. Subsidies could bestructured as scholarships for the poor—perhaps using a village levelmechanism for determining poverty—backed by broader loan-programs tofinance fees for the non-poor (see box 6).

Beyond the crisis, the education system will shape the region’s future workforceand the competitiveness of its economies. Sustaining high quality and broad-based educational expansion is central to equipping workers with the skills forhigh productivity manufacturing and service industries, and to train them overthe course of a working life.

As noted above, Korea has done exceptionally well in this respect, althoughrecently it has been reassessing its education strategy with the goal ofdeveloping more creative and flexible skills. The pressure points in the regionprobably lie elsewhere, for example, in the relative neglect of secondaryeducation in Thailand, upper secondary and college education in China, and inpoor quality education in Indonesia and the Philippines. These issues are onlypartly a question of government spending priorities, as illustrated by Koreawhere secondary, and especially tertiary education is mostly privately financed.Institutional and policy reforms are required to foster the high quality schoolingwhich includes the skills that will propel East Asian countries into theknowledge economy of the next century.

High quality schooling requires reforming curricula at the primary andsecondary levels to emphasize team building, flexibility, and adaptability which


are built on a foundation of literacy and communication skills, plus numericaland analytical skills. To move their economies forward in the early 21stcentury, East Asia's young people will have to master the multitude ofworldwide sources of information and be able to aggregate this knowledge toanalyze and solve local economic and social challenges. Thailand and Malaysiaare working to reform their education system to provide those skills, and Chinais also moving in that direction.

Box 6: Preserving the Poor’s Human Capital during Economic Crisis:Indonesia’s “Back-to-School” Campaign

Impact of the Crisis. In April 1998, focus group discussions and school visitsalready indicated that poor schools and children were feeling the impact of thecrisis. Reduced public funding for education, higher prices of schooling, andlower family incomes are expected to lead to declines in primary and juniorsecondary school enrollments among the poor. Estimating the impact of theeconomic crisis on enrollment is difficult as the crisis is unprecedented in termsof magnitude and depth. Econometric techniques have yielded relatively lowimpact effects ranging from an additional 115,000–260,000 7–12 year olds, and173,000–270,000 13–15 year old children dropping out over time as a result ofa 10 percent fall in per capita income. Estimates from the Government ofIndonesia (GOI) point to much larger effects—an additional 890,000 and640,000 children dropping out of primary and junior secondary schools,respectively, in just one year.

Whatever the precise figures, there is general agreement that the impact of thecrisis on poor children will be severe. The strongest evidence comes from themuch smaller economic shock of 1986–87 when education expendituresdropped and there were no special efforts to keep children in school; grossenrollment rates fell from 62 percent to 52 percent at the junior secondary leveland took almost a decade to recover. Virtually the entire decline came frompoor households. Also, non-salary spending per pupil fell sharply from Rp23,000 (US$18) to Rp 6,000 (US$3) in real terms.

The Response. On July 20, GOI launched a 5-year national program to providescholarships for poor children in basic education and block grants to schoolsserving poor communities. A coalition of Ministers was formed to support theprogram. The World Bank is leading a multi-donor effort, which includes theAsian Development Bank (ADB), UNICEF, and bilateral agencies—AusAID,and Asia Europe Meeting (ASEM)—to support the Program. Total cost of the5-year program is approximately US$382 million, with an ADB contribution ofUS$86 million, and the remainder from the Bank.

Seventeen percent of the poorest students will receive a scholarship of Rp240,000 (US$30) in voucher form at the beginning of the school year. This isintended to cover school costs such as notebooks, uniforms, transportationcosts, and school fees. Nationally, 2.6 million junior secondary students willbenefit (about 17 percent of enrollment). Forty percent of primary and junior


secondary schools serving the poorest communities will receive grants of Rp 2million (US$250) and Rp 4 million (US$500), respectively. A total of 82,000primary and junior secondary schools will benefit from block grants each year.Schools can use the grants to purchase instructional materials and otherteaching–learning supplies, undertake minor repairs, and support poorerstudents by waiving formal and informal school charges.

Mass Media and Social Mobilization Effort. A nationwide TV, radio, andprint media campaign was launched to ensure that parents and communities areaware of the program, to emphasize the importance of remaining in school, andto facilitate transparency in the use of funds and selection of recipients.

Targeting and Selection. Scholarships and grants will be allocated accordingto the poverty incidence of each district. Given the limitation of the quantitativedata, this information will be coupled with local knowledge and theparticipation of NGOs and other members of civil society in the selection at thelocal level. Recipients will be selected by committees at district, sub-district,and school levels, which consist of parents, NGOs, other members of civilsociety, and government representatives.

Ensuring funds reach recipients. In order to ensure scholarships and grantsreach intended beneficiaries, the program includes the following features: (i)funds will go directly from the (local) bank to students–schools—nointermediaries; (ii) a mass media campaign down to the village level will informcommunities and parents of the program and procedures; (iii) an independentagency will carry out quarterly monitoring; (iv) NGOs–civil society memberswill monitor the program; and (v) the Government, the Bank, and the ADB willevaluate the impact of the program on school enrollments and transition throughfocused surveys and the use of SUSENAS.

Health careAnalysis from Indonesia and Malaysia indicates that spending on health centers,and particularly sub-centers, benefits the poor, but spending on hospitals isunequally distributed. There are concerns that price hikes on imported drugslead to postponing or curtailing drug use, including vaccinations and HIV/AIDSdrugs, and that sharp cost increases in private medical services are inducinggreater demand for public services (see box 7). This suggests three measures:

• Preserve real spending on public goods or health activities with highexternalities, such as vaccinations and vector control

• Maintain spending for health centers and sub-centers, especially for non-salary items

• Provide temporary subsidies for essential drugs, during a transitional periodof exchange rate disequilibrium. Such subsidies will likely be weaklytargeted to the poor, but they may still be justified in terms of protectingoverall human resources.


Box 7: The Crisis and Health: A Common Set of Problems

Medical costs are increasing. Exchange rate depreciations have meant largeincreases in medical costs given the high import content of pharmaceuticals,including vaccines and contraceptives. In Indonesia, imports account for 60percent or more of the pharmaceuticals used in the country, and drug priceshave reportedly increased two or three fold. This change in relative prices isunlikely to be fully reversed, and will require long-term adjustments in drugconsumption patterns.

Private consumption expenditure is falling, particularly among the risingnumbers of unemployed. Many households are less able to pay for the out-of-pocket cost of medical care, whether provided by the private sector or by publicsector facilities that typically charge nonzero user fees. This is importantbecause private spending finances 50 percent of aggregate health expendituresin East Asia. There is already evidence that private sector users are switchingback to the subsidized public sector, while some potential users—especiallyamong the poor—may have to switch to lower quality providers, or even foregomedical care entirely.

Public health expenditures are declining. Budgetary pressures can reducepublic subsidies, which protect the poor from the increased financial risks ofillness. This either increases financial hardship, or reduces use of medicalservices. Moreover, increased demand for public services from former users ofprivate facilities could divert public subsidies from the poor. In the long-term,cuts in operations and maintenance outlays will also undermine the productivityof the public infrastructure. Reduced public expenditure also threatens prioritypublic health programs, such as immunization against childhood diseases andTB control. Indonesia’s past experience with fiscal adjustment in the mid-1980sdemonstrates the vulnerability of public health programs to public expenditurecuts.

Institutions, corruption and the social fabric

East Asia’s reputation as a model of reasonably efficient institutions and socialstability has been shattered by the crisis.

• are public sector institutions so riddled with corruption and cronyism thatthey can no longer deliver results?

• is there potential for irreversible social breakdown including rising ethnic orother factional violence, and destruction of community and familybehavioral norms?

Promoting effective institutions


Corruption and poor institutional performance shoulder much of the blame forthe crisis. Corruption is a long-standing feature of most East Asian societies butits profile is on the rise with increasing public attention to internationalcorruption rankings, and high-profile scandals from Japan to Vietnam. In thedecades of extraordinary growth, corruption coexisted with reasonably effectiveinstitutions, from core macroeconomics management bodies to schoolingservices. Now most observers are concerned that public institutions are largelyineffective and driven more by private gain than the public good, especially inIndonesia.

Institutional capacity to deliver resources or services effectively is linked tobroader governance concerns. Indonesia has a reputation for institutionalizedcorruption in which big and small officials stole a share of the public pie.Evidence of institutional weakness is scattered, but it ranges from the lowquality of education, health, and other services to the widespread concerns overgraft and the view that local resource allocations are determined by politicalpower, rather than developmental and social needs. However, East Asia couldnot have enjoyed the massive advances in social conditions if governmentservices—a major part of this effort—were useless. Careful micro studies arescarce, but reveal a mixed picture. A comparison of public irrigation workers inKorea and India found sharply better performance in Korea.38 Recent researchon local institutions in Indonesia finds reasonable performance of publicinstitutions at the village level—although they are significantly worse thangenuine community organizations (preliminary work on the Local Institutionsstudy from World Bank staff and the study team).

Institutional reform is a complex and long-term process and decentralization,though often desirable, is not a panacea especially during the crisis wheneffective delivery is unusually urgent. In fact decentralization of power to localelites can make things worse, in part because they generally have weakertechnical resources at their disposal.39

Evidence from reforms in Latin America and elsewhere has emphasized bothgiving more choice and more voice to the communities who use services.40 Inparticular, a diversified approach to delivery of transfers, involving government,civil society and religious institutions, can help reduce the risks of relying ononly one channel. Restrictions on local associations should also be lifted toencourage entry and competition. This can be complemented by measures tofoster genuine participation of communities in the choice, design,implementation, and evaluation of projects. Social funds were much used inLatin America and Sub-Saharan Africa in response to adjustment. There is alsoincreasing evidence that they perform best when there is genuine participationof local communities.41 Finally, public and independent information, includingmonitoring by civil society organizations, is a potentially valuable source ofincreased accountability.

Responding to a deteriorating social fabricResponses to the added pressure on social relations within the family and thecommunity vary across different societies (see box 8). It is too early to assess


the consequences of social changes, but there are parallels in other communitiesunder economic crisis. Economic decline contributed to rising violence in urbanLatin America in the 1980s. Even though much of the region recoveredeconomically in the 1990s, violence remains high, with widespread economicand social costs. Studies in poor urban communities in Ecuador, Hungary,Zambia, and the Philippines found:42

• Increased work for women and children

• Increased pressures on women and older girls—mothers work more, so girlsmust care for younger children

• Increased substitution of private for public services—including in healthand education

• Increased street violence, especially amongst young males

• Increased domestic violence, especially in households hit by fallingemployment or declining incomes.

Box 8: Erosion in Social Capital

Most citizens see the crisis eroding social capital—trust, reciprocity, andnetworks of support. Nevertheless, in some communities social cohesion mayactually strengthen as poor communities discover resourceful ways to overcometheir problems. For example, in Davao, Mindanao, Philippines the communityinitiated a savings scheme to cover the costs of festivals, and a self-policingprogram (ronda) was developed in response to increased crime.

Conflict. In all countries, NGOs identified increased conflict—within thehousehold, the community and society at large. Increased pressure led to a risein domestic conflict, and “loan sharks” in Bangkok attacked people who couldnot repay their debts to money lenders. NGOs also expressed concern over thepotential for social unrest, a concern which manifested in Indonesia where therehas been extensive ethnic violence against the Chinese population, includingrape.

Vulnerability and insecurity. In Teparak slum settlement in Khon Kaen, innortheast Thailand, a focus group identified a breakdown in community trustwithin the last six months. Increased competition for jobs meant that neighborswho once cooperated were now competing. Theft, violence, and other crimeswere on the rise. Some children had been forced by their parents to drop out ofschool to guard their homes because both parents were now working outsideand break-ins had increased, and neighbors were often the culprits.

Isolation. All focus groups reported a general feeling of uncertainty andisolation. Many said that although the poor had benefited from improved socialwelfare, they still felt excluded and felt that they had not received their fairshare of the economic growth. Many blamed the rich for the current crisis andwere unable to understand why the poor should carry the burden. In Teparak,Khon Kaen, Thailand, a community leader added, “The crisis has happened too


quickly and has left us confused, puzzled, and let down. We have been laid offbut given no explanation.”

Source: Rapid social assessments undertaken by Bank staff in collaborationwith NGOs in Thailand, Philippines, Indonesia and Cambodia, January-April,1998.

The study found varied effects on informal community43 functioning. Moderatepressure could lead to heightened mutual support—increased use of socialcapital—whereas severe pressures were more likely breakdown of community-based coping mechanisms.

These substantial costs underline the importance of restoring the national andlocal economic environment crucial for social functioning. Specific responsescould include:

• Identifying the vulnerable groups and focusing action to respond to theirneeds. For example, heightened pressures on women can be relieved byreducing the demands on their time, such as improving water supply orchildcare facilities.

• Facilitating participatory processes mediated by government, NGOs, orreligious institutions—both to set local priorities and to support informalnetworks.

• Supporting innovative action to reach groups at high risk. For example,concerted action by civil society is often the only way to reach childrenforced into exploitative work. Public action was crucial to reducing childprostitution in the Philippines in some case44. In Brazil local groups usetheater, music, and other forms of community engagement to reach kids onthe streets, diverting them from gangs to more productive forms of socialcapital.

Box 9.The World Bank’s Efforts to Help the Poor

The World Bank is helping governments re-initiate growth to manage the socialconsequences of the crisis; protect public expenditures targeted for the poor;enhance the quality of social services; improve design and financing of socialfunds; strengthen social security systems for the unemployed and the elderly;and address key institutional issues. The Bank’s most important activitiesinclude:

In Thailand, a US$300 million loan for a social investment project willfund job creation for the poor and the unemployed through existing labor-intensive government programs; expand training for the unemployed; supportlow income health insurance schemes, small-scale community projects, andlarger municipal projects; and set up a monitoring system to evaluate the impactof the crisis and of public action on the poor. The loan is expected to create


roughly one million months of jobs and an equivalent amount of training. Also,a national poverty map will be drawn based on a nationwide systematicparticipatory assessment which will be an important input for the policy debateon safety net mechanisms

In Indonesia, the Bank has restructured some of the existing portfolioto redirect savings to support income generation and meet basic needs (aboutUS$320 million). A Structural Adjustment Loan (SAL) of US$1 billionincludes a component to protect the poor through expanded labor-intensivepublic works programs; actions to ensure the continued availability of keygoods with only modest price increases; and initiatives to maintain access toquality basic education and health. In particular, to ensure continued highenrollment rates for children through the first nine years of school, theGovernment is to provide scholarship funds for 2.6 million needy juniorsecondary school students. A US$275 million poverty project for rural areas(the Kecamatan Development Project) has been approved, and a similar projectis under preparation for the urban poor. Discussions are also underwayregarding an agriculture sector adjustment loan to support reforms. The Bankhas also intensified analytical and participatory work on poverty to helpunderpin these two operations and will help finance a follow-up to the IndonesiaFamily Life Survey (IFLS). This will allow monitoring of the living standardsof a sub-sample of households that were already surveyed in 1993 and 1997and help assess household-level coping strategies in response to the crisis.

In Korea, the US$2 billion SAL approved in March 1998 includes animportant program on labor markets and social safety nets. The programincorporates measures to increase flexibility in the labor market while extendingcoverage of unemployment insurance to employees in small-scale enterprises;improve poverty monitoring and protect poverty-related public expenditures;and reform the pension system. A conference in July 1998 focused on lessons ofinternational experience in labor market policies. Another conference will focuson pension fund investment policies. A second SAL for US$2 billion will helpdeepen these reforms and start addressing issues in health financing and healthcare.

In the Philippines, three new loans (US$79 million) all aiming toincrease the incomes of the poor and to provide basic services were approved inMarch 1998. Two other loans totaling US$130 million are scheduled forapproval in the second half of 1998; these projects finance infrastructuredevelopment and will increase job opportunities and access to basic services forrelatively poor local government units. The Bank has carried out a rapid socialassessment to gauge the effects of the crisis and understand household copingstrategies. Poverty work scheduled for fiscal 1999 will have access to theresults of the 1998 Income and Expenditure Survey (FIES is carried out everythree years) and contribute to the early implementation of the Annual PovertyIncidence Survey. This will provide a useful analysis of the short-term impactof the crisis, evaluate the effectiveness of government policies to alleviatepoverty, and provide policy directions for the future.

In Malaysia, a US$300 million Economic and Social Sector Loanapproved in June 1998 will support a reduction in the fiscal surplus from 2.5 to


0.5 percent of GDP by increasing public expenditures for the social sectors.The Loan seeks to protect budgetary spending for education, health, ruralinfrastructure, and to increase expenditures on social safety net programs aimedat providing direct support to the poor (free housing and food supplements) andincome generation through small grants. Longer-term issues about the adequacyof formal safety nets and the governance structure of the Employee ProvidentFund will be addressed through a Technical Assistance Loan and economic andsector work. In addition, a CEM is nearing completion—the first since 1993. Itincludes an overview of poverty and the social safety net in Malaysia, analyzeshow the poor may be affected by the downturn, and recommends action tocushion the impact of the crisis on the poor.

In China, ongoing work in labor market adjustment focuses on policiesneeded to address the un(der)employment problem. While the problem isbecoming acute largely due to accelerated reforms in the state-owned enterprisesector, the slowdown in aggregate demand, which will be exacerbated by theimpact of the regional crisis, is also having an impact. A workshop will discussthe effectiveness of active and passive labor market policies in addressingemployment problems.

In Cambodia, a study is underway to examine the impact of theregional crisis on Cambodia and Lao People’s Democratic Republic, includingan analysis of the social impact, in particular through rapid social assessments.

A region-wide initiative is being launched to analyze issues in pensionspolicy and administration. A November 1998 conference will bring togethercountries of the region to explore common issues and frame an agenda forfuture work. This would be followed by a conference in the spring of 1999 ongovernance of pension funds in East Asia.

The central role of public informationThis paper has emphasized the importance of understanding the scope andnature of the crisis. A crucial need is to deepen understanding of how the crisisis affecting households and how different actions are ameliorating the socialcosts. This matters technically because there is much that is not known aboutthe impacts. It matters in terms of governance, as informed public debate aboutthe choices being made should be guided by up to date information. Effortsshould provide regular updates on overall welfare and vulnerability to provideaccountability for how resources are used. Assessing public actions will provideinformation which will be included in ongoing redesign of these programs.Public information will provide checks on transfers and underpin an informedpublic debate on developments, effects of programs, and tradeoffs.

Programs that monitor overall welfare should rely on a mix of instruments:

• Regular surveys of living conditions and vulnerability covering wages,unemployment, relative prices, food prices, drought effects, socialindicators, and nutritional levels. In many East Asian countries, such short-


term information is weak. The Philippines has a Social Weather Stationsystem based largely on subjective assessments. Indonesia recently droppedthe quarterly version of its labor force survey because there was littleseasonal variation. Thailand still has a quarterly labor survey and Korea’slabor survey now takes place monthly.

• Complementary assessments of household and commentary conditionsusing participatory techniques which substantially enrich the understandingof coping strategies. These assessments should form part of the ongoinginteraction between hypotheses and data, and could influence the design ofsurvey questionnaires.

• Use of existing data sets to match economy-wide trends to the structure ofincome and spending, and to analyze past household responses to changesin specific parameters, such as price elasticities of schooling costs. In mostcountries, existing data includes consumption or income surveys.

Monitoring and evaluation of public action are important short-term goals toensure that intended effects reach targeted groups, and to redesign programs. Inthe medium to long term monitoring and evaluation help to bring poor regionsand groups into the development process. All programs can benefit fromcombined quantitative and participatory monitoring. Participatory monitoring isparticularly valuable for increasing effectiveness by strengthening communityinvolvement, and increasing efficiency by scrutinizing the use and allocation offunds. Public information strengthens accountability. For key programs,especially those with uncertain impacts, structured evaluation that uses samplesof participating households and controls is also important for maximizing thebenefits of scarce resources. For example, structured evaluation is likely tocover the targeting efficiency of public works, the incidence of new subsidiesfor education, and the effectiveness of geographic targeting to alleviate poverty.


Conclusion

What began as a currency crisis in Thailand has evolved into a social crisisacross the region and beyond. Within East Asia it was initially hoped that thecrisis would involve a sharp contraction and sharp recovery—a “V”-shapedresponse to a shock, as occurred in Mexico after the 1994/95 currency crisis.That alone would have been an unusually severe shock to East Asianhouseholds, just as the tequila crisis was a major source of distress to manyMexican households. There are now fears that the crisis will at best be “U”-shaped, with a more prolonged period of contraction, and risks of a lengthyperiod of working through of structural, social and institutional problems. Thisis a picturesque way of presenting the tough message that there are no magicbullets or easy solutions to what has become the most serious economic andsocial crisis in decades.

The crisis has been correctly characterized as a regional crisis, but thedevelopment of policies and programs require a recognition that there arediverse issues due to country and localized differences. The disaggregation ofsocial impacts is vital for determining the types of responses to be funded, andhow programs should adapt as local conditions change. This applies, of course,with even greater force to assessment of the social consequences of crises inother regions of the world. This requires that donors expand their relationshipswith civil society organizations both for policy dialogue and for research onsocial impacts.

While it is important to design responses in the context of diverse localconditions, some general conclusions can be drawn. In the midst of the crisis,there are four domains where a combination of immediate action andexploration of alternatives is central:

• The distributional effects of both the shocks and economy-wide responses,between rich and poor, across different socio-economic groups (withparticular attention to socially excluded groups), and between men, womenand children. This remains both a powerful and ill-understood domain.

• The potential for direct public action, especially using public resources, toboth reduce income losses due to employment declines and priceincreases—through various forms of safety nets—and reduce the risk ofirreversible losses in human capability through lost education, malnutritionand foregone action in health. Here there is a relatively deep understandingof the policy options and their likely effects, but there are sometimesformidable barriers in their execution, in the context of aggregate fiscalretrenchment (and associated distributional struggles) and institutionalweaknesses.

• The consequences for the social fabric of economic decline and cutbacks insocial services, especially when this is in the context of societies in rapidand turbulent transition from traditional to “modern” social structures.Case-studies have provided increasing evidence of adverse social effects,


from rising violence, to increased prostitution and intensifying pressures onwomen. But public action to offset these social consequences is rarelyundertaken in a systematic fashion.

• The monitoring and evaluation of the effects of the both economy-wide andlocal developments and various forms of public action on individual andcommunity welfare, through both structured quantitative and participatorytechniques.

• In previous crises involving the international community, including theWorld Bank, the primary focus has been on the second domain above, andeven this is weakly understood and documented. Action in the otherdomains has typically been dismal. It is essential that the domestic andinternational community does better in this crisis.

Confronting the immediate social fallout of the crisis is crucial, but it is of equalimportance to shape responses in the context of longer-term developmentchallenges. Some may see the crisis as providing the basis for a fundamentalquestioning of the model of economic and social development based oninternational engagement and reliance on markets. This would be mistake. EastAsia’s success prior to the crisis was not a mirage, and a high proportion of theassociated social gains will still be there after the crisis has resolved. However,the fact that East Asia was both a paradigm of success and then the epicenter ofthe current crisis illustrates a crucial point. The forces of globalization andtechnological advance provide the potential for an unprecedented pace of socialadvance. But these same forces also carry risks—of pressures for risinginequality between those groups equipped to participate in today’s globaleconomy and those not so equipped (for lack of skills, lack of access tocomplementary domestic or foreign capital, or through exclusionary socialforces); and of rising risks of income and social losses, at an individual,community, national and regional level. The key challenge for social andeconomic policy at the turn of millennium, is understanding and shaping thepolicies, institutions and social structures that will help realize the potentialwhilst managing the risks. Just as there are issues of international financialmanagement to mediate these competing pressures, so there are questions offostering structures that foster inclusive and equitable social and economicdevelopment and provide mechanisms for social protection consistent withsocieties that are both caring and competitive.

A particular aspect of the lessons from the Latin American experience that hasreceived little attention in many studies is the relationship between politicalstructures and effective reforms. Some of the contributing factors to the crisis interms of lack of transparency, corruption, cronyism, etc. point to the need for afundamental opening of decision making in economic policies. It is in part inrecognition of this, that the World Bank is encouraging engagement with civilsociety organizations needs to be deepened and expanded at the center ofaddressing the social and economic costs. This includes project implementationand research on social costs, but is wider and deeper in terms of finding newways of bringing diverse parts of civil society into determining responses to thecrisis.


In order to better address the social questions, and acknowledging that allorganizations involved in the crisis countries need to work together, the WorldBank has set up a number of initiatives to support dialogues, informationsharing and engagement with civil society. The bulk of this is occurring at acountry level, but there are also regional initiatives to foster broader debate andexchange of information. These include a website, moderated dialogues that willbe reported on the website, regular country consultations with civil society,various forms of regional consultations and new research partnerships.

In the midst of the crisis, immediate action is required. In this paper we make arange of suggestions where public action can reduce—though almost certainlynot fully offset—the severe human costs of an economic crisis that started incurrency markets, stock exchanges and banks. However, the paper has alsoplaced considerable emphasis on the inadequacies of both past social responsesto crises and of our present understanding of how to do better. This paper hasfocused on East Asia, but just as the financial crisis has spread to other regions,so all the issues presented here apply to other countries and regions as well. Anopen process of exploration and exchange is of central importance both for theeffective design and implementation of any response, and for the continuingexploration of how to deepen our understanding of the personal and societaleffects of crisis, of the options for action, and how to reduce the human toll offuture shocks and ensure an inclusive pattern of longer-term development whenEast Asia emerges from the present troubled times.


Bibliography

Agrawal, Nisha and Michael Walton. 1996. Women at Work in East Asia. DoesStellar Growth Meet the Needs of Women? Should Governments do More?Washington, D.C.: World Bank

Ahuja, Vinod and Deon Filmer. 1996. “Educational Attainment in DevelopingCountries: New Estimates and Projections Disaggregated by Gender”. Journalof Educational Planning and Administration. 3:229-54.

Ahuja, Vinod. Benu Bidani, Francisco Ferraira, and Michael Walton. 1997.Everyone’s Miracle? Revisiting Poverty and Inequality in East Asia. ADirections in Development book. Washington D. C.: World Bank.

Aoki, Masahiro, Hyung-Ki, Kim and Masahiro Okuno-Fujiwara, eds. 1996. TheRole of Government in East Asian Economic Development: ComparativeInstitutional Analysis. New York: Oxford University Press.

Birdsall, Nancy and Richard Sabot. 1993. “Virtuous Circles; Human CapitalGrowth and Equity in East Asia.” Background paper for The East AsianMiracle. Washington, D. C.: World Bank.

Blejer, Mario I. and Isabel Guerrero. 1991. “The Impact of MacroeconomicPolicies on Income Distribution: An Empirical Study of the Philippines.”Review of Economics and Statistics.

Deininger, Klaus and Lyn Squire. 1996. “A New Data Set Measuring IncomeInequality.” The World Bank Economic Review. 10(3):565-91.

Diwan, Ishac and Michael Walton. 1997. “How International Exchange,Technology, and Institutions Affect Workers: An Introduction.” The WorldBank Economic Review. 11(1): 1-15

Ferreira, Francisco H. G., and Julie A. Litchfield. Forthcoming. “Education orInflation? The Roles of Structural Factors and Macroeconomic Instability inExplaining Brazilian Inequality in the 1980s.” LSE-STICERD DiscussionPaper. London School of Economics.

Filmer, Deon and Lant Pritchett. 1997. “Child Mortality and Public Spending onHealth” Policy Research Working Paper 1864. Washington D.C.: WorldBank, Development Research Group.

Graham, Carol. 1998. Private Markets for Pubic Goods: Raising the Stakes inEconomic Reform, Brookings Institution Press, Washington D.C., 1998.

Hammer, Jeffrey S., James Cercone and Ijaz Nabi. 1995. “Distributional Effects ofSocial Sector Expenditures in Malaysia, 1974 to 1989”. In Dominique van deWalle and Kimberly Nead, eds. Public Spending and the Poor: Theory andEvidence. Baltimore, Md: The Johns Hopkins University Press.


Jalan, Jyotsna and Martin Ravallion. 1997. “Spatial Poverty Traps?” PolicyResearch Working Paper 1862. Washington, D.C.: World Bank, DevelopmentResearch Group.

Kim, Dae-Il, and Robert H. Topel. 1995. “Labor Markets and Economic Growth:Lessons from Korea’s Industrialization, 1970-1990.” In Richard B. Freemanand Lawrence F. Katz, eds., Differences and Changes in Wage Structures.Chicago: University of Chicago Press.

Mazumdar, Dipak and Zafiris Tzannatos. 1998. “Industrial Development and LaborMarket Segmentation in East Asia: A Comparative Analysis”. Washington,D.C.: World Bank, Human Development Department.

Narayan, Deepa and Katrinka Ebbe. 1997. “Design of Social Funds. Participation,Demand Orientation, and Local Organizational Capacity”. World BankDiscussion Paper No. 375. Washington, D.C.: World Bank.

Pencavel, John H. 1995. “The Role of Labor Unions in Fostering EconomicDevelopment.” Policy Research Working Paper 1469. Washington, D.C.:World Bank.

Radhakrishna, R. and K. Subbarao. 1997. India’s Public Distribution System: ANational and International Perspective. World Bank Discussion Paper 380.Washington, D.C.: World Bank.

Ranis, Gustave. 1996. “Another Look at the East Asian Miracle. The World BankEconomic Review 9:509-34.

Ravallion, Martin. 1998. “Appraising Workfare Programs.” Washington, D.C.:World Bank, Development Research Group.

Ravallion, Martin and Shaohua Chen (1998a). “A Note on the Measurement ofIncome Inequality in Post-Reform Rural China”. Washington, D.C.: WorldBank, Development Research Group.

___________. 1998b. “Poverty Rates in East Asia with Zero Growth.” Mimeo.

Robb, Caroline M.,1998 “Social Impact of the East Asian Crisis: Perceptions from

Poor Communities.” Paper prepared for the East Asian Crisis Workshop,IDS, University of Sussex, UK, June 1998.

Subbarao, K., A. Bonnerjee, J. Braithwaite, S. Carvalho, K. Ezemenari, C. Graham,and A. Thompson. 1997. Safety Net Programs and Poverty Reduction:Lessons from Cross-Country Experience. A Direction in Development book.Washington, D.C.: World Bank.

Teranishi , Juro. 1996. “Sectoral Resource Transfer, Conflict and Macrostability inEconomic Development: A Comparative Analysis.” In Masahiko Aoki,


Hyung-Kim, and Masahiro Okino-Fujiwara, eds., The Role of Government inEast Asian Economic Development: Comparative Institutional Analysis.New York: Oxford University Press.

UNESCO. 1997. UNESCRO Statistical Yearbook. Paris: UNESCO Publishing.

United Nations. 1995. World Urbanization Prospects: The 1994 Revision. UnitedNations, New York: New York.

van de Walle, Dominique. 1995. “The Distribution of Subsidies through PublicHealth Services in Indonesia, 1978-87.” In Dominique van de Walle andKimberly Nead, eds., Public Spending and the Poor: Theory and Evidence.Baltimore, Md: The Johns Hopkins University Press.

Wade, Robert. 1994. “The Governance of Infrastructure: Organizational Issues inthe Operation and Maintenance of Irrigation Canals.” World Bank.

World Bank. 1990. World Development Report 1990: Poverty. New York: OxfordUniversity Press.

__________. 1993a. The East Asian Miracle: Economic Growth and PublicPolicy. A Policy Research Report. New York: Oxford University Press.

__________. 1993b. “Indonesia: Public Expenditures, Prices and the Poor.” Report11293-IND. Washington, D.C.

__________. 1994. Averting the Old Age Crisis: Policies to Protect the Old andPromote Growth. World Bank Policy Research Report. Oxford UniversityPress for the World Bank: Washington, D.C.

__________. 1995. World Development Report, Involving Workers in East Asia’sGrowth. New York: Oxford University Press.

__________. 1996a. Involving Workers in East Asia’s Growth. RegionalPerspectives on World Development Report 1995. Washington, D.C.

__________. 1996b. World Development Report 1996: From Plan to Market. NewYork: Oxford University Press.

__________. 1996c “Confronting Crisis. A Summary of Household Responses toPoverty and Vulnerability in Four Poor Urban Communities," ESD Studiesand Monograph Series No. 7, 1996.

__________. 1997a. China 2020: Development Challenges in the New Century.China 2020 Series. Washington, D.C.

__________. 1997b. Sharing Rising Incomes: Disparities in China. China 2020Series. Washington, D.C.


__________. 1998a East Asia: the Road to Recovery. Washington, D.C.

__________. 1998b (forthcoming) Global Economic Prospects. Washington, D.C.

__________. 1998c (draft) The Poor in Indonesia’s Crisis. Washington, D.C.


END NOTES

1 World Bank (1998a).2 See World Bank (1998b) [Global Economic Prospects—forthcoming].3 Many groups in civil society and some researchers were all too aware of the socialeffects of the shocks and responses; amongst international agencies, UNICEF, inparticular, was ahead in emphasizing their importance.4 This section draws on Ahuja, and others. (1997).5 World Bank (1993a); Birdsall and Sabot (1993); Teranishi in Aoki, and others.(1996).6 World Bank (1995); World Bank (1996a).7 See Kim and Topel (1995).8 See Ranis (1995).9 After controlling for incomes and other characteristics (though in Indonesia thismay partly reflect unusually rapid income growth, and the slower response ofmortality. See Filmer and Pritchett (1997): Indonesia is an outlier (that is, had highchild mortality) after controlling for incomes; the Philippines and Korea werenegative outliers after controlling for incomes and a set of other characteristics,including female education and inequality.10 Ahuja and others, (1997).11 Jalan and Ravallion (1998).12These comparisons ignore issues such as equivalence scales and differences inregional cost of living. In addition, the discussion includes Ginis for both expenditureand income distribution. Since income distributions are generally more unequal thanexpenditure distributions, comparisons are only valid across indices for the sameconcept (for example, over time). Similarly, we include distributions per householdand per individual, which, once again, are not strictly comparable. The figuresmerely indicate trends.13 Preliminary analysis of the 1997 survey yields an increase from 45.1 in 1994 to49.6 in 1997 in the Gini coefficient for family income. The absolute numbers are notto be compared to the data cited in table 4 which are Gini coefficients for expenditureper capita but the trends and magnitude of changes are likely to be similar.14 It is possible to observe increases in inequality that leave poverty unchanged for agiven level of average income. A mean-preserving spread originating from a transferfrom an individual above the poverty line will increase inequality but not affectpoverty.15 See Mazumdar and Tzannatos (1998).16 World Bank (1996a).17 Based on The Social Weather Station surveys of Filipino’s perceptions of theirpoverty.18 See Agrawal and Walton (1996).19 Robb (1998).20 See World Bank (1990).21 Projections are from Ravallion and Chen (1998). The methodology involvedupdating household data to 1997 using actual or estimated growth in averageconsumption or income per capita, but assuming no distributional changes since themost recent survey—survey years are reasonably recent: Indonesia (1996), Malaysia(1995), Philippines (1994), and Thailand (1992). Then a range of projections for the


entire distribution was developed by assuming alternative values for growth in themean level of consumption or income and the degree of inequality. The change ininequality was estimated in terms of the shifts in the parameters of the Lorenzdistribution to produce a certain percentage change in the Gini coefficient.22 This scenario was developed by Benu Bidani, as part of ongoing work on povertyin Indonesia and is discussed in “The Poor in Indonesia’s Crisis, (mimeo),” WorldBank (1998).23 “The Poor in Indonesia’s Crisis” (draft). World Bank. (1998).24 See World Bank (1990), Chapter 7.25 See Sarel (1996): that actually found rising investment efficiency in Indonesia inthe past decade.26 Work on the U.S. finds that unemployment disproportionately hurts the poor(Blinder and Blank). Similar effects may explain some of the recession-linkedincreases in inequality in Latin America.27 Blejer and Guerrero (1991); Ferreira and Litchfield (1997).28 Banking with the Poor Network Newsletter, Issue No. 11, June 1998, pp1–8.Foundation for Development Cooperation, Brisbane.29 See Wood, Pissarides and Tan, World Bank (1997a).30 See Staff Appraisal Report for South West Poverty Reduction Project (May1995) and Qinba Mountains Poverty Reduction Project (May 1997)31 This is because per capita rice consumption does not vary much acrossexpenditures classes (the bottom 15 percent of the population consumes 13 percent ofthe rice). Therefore the proportion of the rice subsidy that goes to the poor (and theleakage to the non-poor) is no different from that of a general cash transfer.32 In some poorly targeted Indian programs 6–7 rupees are spent to transfer onerupee to a poor household. Radhakrishna and Subbarao (1997).33 See Ravallion (1998), “Appraising Workfare Programs.”34 See Subbarao and others. (1997).35 Pencavel (1995) and World Bank (1996a).36 World Bank (1996a).37 See World Bank (1993).38 See Wade (1994).39 World Bank (1997b).40 See Graham (1998).41 See Narayan and Ebbe (1997).42 World Bank (1996b).43 See World Bank (1995).44 See World Bank (1996b).

Documents

Social Consequences of the East Asian Financial Crisis€¦ · Social Consequences of the East Asian Financial Crisis. 3. East Asia’s economic and social structures are under strain,